Home

EO-1/ Hyperion Science Data User's Guide, Level 1_B

1. Li 1 4 Figure 11 VNIR SVVIR Spatial Co registration The Level 1 data set consists of radiometrically corrected images formatted as HDF files and metadata in binary and ASCII formats The data format and units for the data files have changed as the level 1 processing has been modified For data processed using the Level 1 code the data type is 16 bit unsigned integer and the units are vvatts sr micron m x 100 For data processed using the Level 1 A or Level 1 B code data type is a 16 bit signed integer with units of watts sr 32 EO 1 Hyperion Science Data User s Guide HYP TO 01 077Rev Public Release Ll B micron m x40 for VNIR data and watts sr micron m x80 for SWIR data Both the file extension and the header can be used to determine the version of the level 1 processing code Table 8 defines all Level 1 Level 1_A and Level 1_B data products Level 1 metadata
2. Source file File name drbl2 archive levell ground E012000147 O1CA01C9 r1l smear Type Hyperion Dimensions 256 pixels x 242 bands x 206 frames Data type 16 bit unsigned integer Byte order big Pixel order BIL Output file File name drb12 archive levell ground E012000147_01CA01C9_rl echo Type Hyperion Dimensions 256 pixels x 242 bands x 206 frames Data type 16 bit unsigned integer Byte order big Pixel order BIL Source drb12 archive levell ground E012000147_01CA01C9_rl smear Ratio file Ratio file dra2 calfiles ratio txt Start band 71 KKKKKKKKKKKKKKKKKKKK Figure 16 Example MD 8 Echo Removal Log 41 EO 1I Hyperion Science Data User s Guide HYP TO 01 077Rev Public Release L1_B The Smear Removal Log MD9 has the same format as the Echo Removal Log MD8 but there is no file used in the smear correction process Again the band at which smear correction begins is reported The VNIR focal plane covers spectral bands 1 70 so it is expected that smear correction which is also only required in the SWIR would start at band 71 MD9 EO12000147_01CD01CC_r1 smear log KKKKK hypsmear KKKKK HIP 1 1 Mon Jun 5 13 07 31 2000 Arguments source file drb12 archive level0 ground E012000147_01CD01CC_r1 L0 source type hyperion output file drb12 archive levell ground E012000147_01CD01CC_r1 smear output type hyperion start band 71 verbose Source file File name drbl2 arch
3. 660 lines corrected files gt Average frames of all gt Repair known bad pixels gt Display image in ENVI gt Both dark and image files echo removed dark files gt Log file generated MD10 gt Manual evaluation eae gt Log file generated MD8 MD3 spectral profiles animation MD11 gt Generate ENVI header file MD12 Add text file containing pre flight characterization MD1 Create tape and ship to GSFC for distribution to Science Team 34 EO 1 Hyperion Science Data User s Guide HYP TO 01 077Rev Public Release L1_B Figure 13 Level 1 A Data Processing Flow Diagram Receive LO data from GSFC Dark cal interpolation gt Linearly interpolate between preimgdk avg MD3A and postimgdk avg files MD3B SEC Flag pixels gt 4095 gt Pre amp post image darks gt Image gt Output to log file MD15 Step 4 Apply calibration gt Multiply dark subtracted Background removal image by lab gain file MD7 to gt Subtract interpolated dark obtain radiometrically corrected file from echo and smear Level 1 data corrected image file gt Multiply VNIR radiance by 40 gt Multiply SWIR radiance by 80 gt Log file generated MD2 Smear correction gt Apply smear correction to SWIR data gt Both dark and image files gt Log file generated MD9 Average dark files gt Average frames of echo and smear corrected pre Fixstripes Co registration of VNIR SWIR and post image dark files
4. Hyperion VNIR 5VA 5 02V Hyperion VNIR 5VA 4 94V Hyperion VNIR 15VA 14 98V Hyperion VNIR 15VA 14 92V Hyperion SWIR 5VD 4 99V Hyperion SWIR 5VA 5 03V Hyperion SWIR 5VA 4 95V Hyperion SWIR 15 Volts 14 99V Hyperion SWIR 15 Volts 14 9V Cal lamp power supply voltage 13 0 when on Hyperion command counter variable Hyperion error counter variable Cryocooler coldhead temperature 110K when cryocooler is operating Cryocooler 10 Log10 RSS of all harmonics Cryocooler motor drive 75 89 when cryocooler is operating 3 4 4 Explanation of SDS Attributes in Level 0 product The SDS attributes for the Hyperion Level 0 data files are listed in Table 7 A description of each of the attributes follows The values in the example represent a pre image dark collect This level of detail is provided for the advanced user Table 7 SDS Attributes for Hyperion Level 0 Data NAME Max HDF Type Example Value Length Image Attribute 256 8 bit character Pre image dark Level 0 File Generated By 256 8 bit HLZP version 1 0 0 character Byte Order 256 8 bit big character 27 EO 1 Hyperion Science Data User s Guide HYP TO 01 077Rev Public Release Ll B NAME Max HDF Type Example Value Length Level 0 File Generated At 14 8 bit character yyyy_ddd_hhmmss Frame Numbers 2 32 bit unsign
5. L1 _A3 Saturated pixels A modified version of cubesat repaired errors in reported sample number of saturated pixels removed smear and echo affected pixels for VNIR bands 1 70 from the sat file and sorts the sat file by band number L1_B Offset removal Eren ere The dark file interpolation algorithm was modified to account for the time between the pre image dark file and the start of the Eren ere image Saturated ac The saturated pixel report was updated to reflect the spatial shift of SWIR pixel locations as a result of VNIR SWIR co pene ae The new report includes saturated pixels locations for the LO and L1_B data VNIR SWIR co Level 1_B co registers the VNIR and SWIR data Only SWIR data are shifted into alignment with the VNIR pixels In the FOV alignment dimension the SWIR pixels are shifted by 1 In the along track dimension FOV pixels 129 256 are shifted 1 pixel A new metadata file is generated with a aln log extension 56 EO 1I Hyperion Science Data User s Guide HYP TO 01 077Rev Public Release L1_B Ratio txt File Modifications zeroed pixels ratio_revB txt Band Sample 57 EO 1I Hyperion Science Data User s Guide HYP TO 01 077Rev Public Release Ll B Bad Pixel File History bad pixel file history 93 Ld Se ea badpix3 er Tere 1 1 dead 1 1 dead 2 1 dead 2 1 dead
6. Smear Removal Log sssessesesesesseeseseresressessresressessresesseeseerreeseesee 42 Figure 18 Example MD 10 Pixel Repair Lope ee die iii te cecs duel 44 Figure 19 Example MD 17 VNIR SWIR Co registration Log ceeeeeceeseeeeseeeneeceteeneeeeenees 45 Figure 20 Image of the Center Wavelength Calibration File oo cece eesecsseceseceseeeeneeenaeeneensees 48 Figure 21 Image of the Full Width Half Maximum Calibration File eee eeeeeseeesteeeseeteeee 49 Figure 22 Variation of the VNIR center wavelength across the field Of view 50 Figure 23 Variation of the SWIR center wavelength across the field Of view secsec 50 EO 1 Hyperion Science Data User s Guide HYP TO 01 077Rev Public Release Ll B Table 1 Table 2 Table 3 Table 4 Table 5 Table 6 Table 7 Table 8 Table 9 List of Tables Radiometric Performance iii ace er ieee eileen ee 11 Special PST Ron aCe er ti ede ecg sh ae amon ctr astue a soup adr aw anaes tueodna er eeoaams overs oantee 11 Image Quality Performances eke Bice Sree Meee ee Bla EE 12 B sic Data Collection Event Timelines 2ccccssccise acs dindi eect eae Rien enieee 23 Types of Data Collection Events emeten pensat ne e tea ue 24 Hyperion Mnemonics included in the Ancillary Data e oo22200222 022 0nuaaaneroneeaes 26 SDS Attributes for Hyperion Level 0 Dat zsmienio mai sat saoves sndaecesndadsheaseeasdevantaavte 27 Hyperion Level l Data Sets aneso ae t airen e
7. The MET is also recorded with the image date To convert to Greenwich Mean Time the Universal Time Correlation Factor 567 648 000 sec must be added to the MET Table 6 Hyperion Mnemonics included in the Ancillary Data Mnemonic Description typical values YIMAGE MODE Image mode idle standby imaging YCOVRSTAT Commanded cover position status O closed 1 solar cal position 2 open YLAMPIVAL Calibration lamp 1 commanded value 0 26 EO 1 Hyperion Science Data User s Guide HYP TO 01 077Rev Public Release Ll B Mnemonic YLAMP2VAL YLAMPICUR YLAMP2CUR YLAMPIVOL YLAMP2VOL YHSATEMPS YHSATEMP6 YCOVERPOS YSWIRFPET YVNIRFPET YHEAPSV YHEAPISV YHEANISV YVNIRPSVD YVNIRPSVA YVNIRMSVA YVNIRPISVA YVNIRMISVA YSVVIRPSVD YSWIRPSVA YSWIRMSVA YSWIRPI5V YSWIRMI5V YCLAMP12VOL YCOMCTR YERRCTR YCOLDHEADTEMP YOUTDBRSS YMOTORDRIVE Description typical values Calibration lamp 2 commanded value 0 255 Hyperion lamp currents 0 98A when on note YLAMPICUR is always 0 Hyperion lamp voltages 7 4V when on note YLAMP1VOL is invalid VNIR ASP temperature 32 34C SWIR ASP temperature 4 5C Hyperion cover position counts 3424 closed 3212 solar cal 2678 full open Hyperion SWIR FPE temperature 153 5 1C when at proper operational temperature Hyperion VNIR FPE temperature 2 10C Hyperion HEA 5 volts 4 8V Hyperion HEA 15 volts 14 99V Hyperion HEA 15 volts 14 9V Hyperion VNIR 5VD 5 22V
8. The first step in both the Level 1_A and Level 1_B processing codes is to identify saturated pixels The first step in the originally released Level 1 processing code was to correct for image artifacts This is the second step in the Level 1 A and Level 1_B codes There are two artifacts in the SWIR echo and smear Because the corrections for echo and smear depend on signal level corrections based on a saturated signal level are invalid Level 1_A produces a metadata file MD15 that reports all saturated pixels as well as those pixels whose derived smear and echo corrections will be invalid MD15 is slightly different in the Level 1_B code in that it reports the location of saturated pixels before LO and after co registration L1_B Following the flagging step in Level 1 A and Level 1 B codes all Level O SWIR data image and dark are corrected for smear and echo artifacts The next step is to subtract off a dark frame This is required because each image includes not only the scene signal but also a signal caused by thermally generated electrons in the bulk material To enable removal of this signal from the image a pre image and a post image dark frame are taken as part of each DCE Each dark file is 1 second of data corresponding to approximately 220 frames The original Level 1 processing code performed the dark subtraction using the averaged dark file that was obtained closest in time with the image The Level 1_A and Level 1_B process
9. images The flow of the data from the user request to the shipment of the data to the user is highlighted Then the details of the level 0 and level 1 processing are discussed For the level 1 22 EO 1 Hyperion Science Data User s Guide HYP TO 01 077Rev Public Release Ll B processing discussion the steps as well as the data files that are created at each step of the data processing sequence are presented with sample metadata files described as examples 341 Hyperion Data Collection Event Sequence The basic Hyperion Data Collection Event DCE consists of three dark collects the scene and a lamp collect A sample timing of the sequence and the naming of each collection are provided in Table 4 below The times are presented in terms of mm ss minutes seconds relative to the specified scene collection time For example the instrument is placed in Standby mode 10 minutes 31 seconds before the start of the image collection The instrument is placed in Idle mode 4 minutes after the end of the image collection Although this format may seem unnatural it is the form the EO 1 planners use in scheduling Hyperion collects and is used here for consistency Hyperion collects data at a frame rate of 223 4Hz The dark collection is 1 second 220 frames the image is typically 30 sec 6925 frames and the lamp collect is 3 seconds 660 frames The output filename is in this format EOlyyyyddd_vvvvssss_rl_ggg XX_ LO This format is described in se
10. 4 1 2 Verification of the Absolute Radiometric Calibration The absolute performance verification plan cross checked radiance measurements from three different paths solar calibration internal lamp calibration source and Lake Frome vicarious calibration effort For the solar calibration comparison the sun s irradiance based on the Hyperion measured radiance was compared with solar irradiance models in the literature The agreement was within 2 in the VNIR and 5 8 in the SWIR The in flight calibration lamp was used as a ground to on orbit transition The lamp results indicated the change in responsivity was less than 3 in the VNIR and 5 8 in the SWIR Vicarious calibration using Lake Frome in Australia was incorporated into the performance verification of the Hyperion imaging spectrometer instrument The ground reflectance measurements and atmospheric correction leading to Top of the Atmosphere TOA radiances are consistent with the Hyperion ground and solar calibration at the 5 level in the 450 to 850 nm spectral range The SWIR agreement is 10 to 15 The efforts generally indicated the VNIR was within the accuracy of the measurement The SWIR was within the accuracy of the comparisons However the SWIR results consistently indicated a lower response 4 2 Radiometric Calibration Advanced Topics Discussions of the SWIR operational temperature pixel to pixel improvements and optical scatter is provided for the more advance
11. DCE downlink to level 1 processed data typically varies from 1 5 to 3 weeks depending on the ground station used because of the length of time it takes to get data tapes from the ground station to GSFC using surface mail 1 6 Hyperion Contact List The following contact information is provided for Hyperion related questions Debra Beiso 310 812 5244 Hyperion Performance Analysis and Data debra beiso trw com Processing Mr Steve Carman 310 812 0279 Hyperion Program Manager steve carman trw com Dr Jay Pearlman 310 812 0337 EO 1 Program Support Science and jay pearlman trw com Mission Operations Dr Carol Segal 310 813 0229 Hyperion Deputy Program Manager carol segal trw com Mission Operations and Planning There are many supporting documents for the Hyperion instruments This includes memos addressing instrument performance as well as papers prepared for journal and conference publication The user should contact Carol Segal for requests of supporting documentation 2 HYPERION DATA CUBE QUICK START This chapter is designed to provide the user with some quick steps to get familiar with the data The discussion assumes that the user is using the Hyperion Level 1 processed data The file should be the Level 1 radiometrically calibrated data with a L1_B extension Data processed prior to July 1 2001 will have a L1 extension data processed prior to November 15 2001 will have a L1_A extension This description is geared towards users
12. SWIR Channel Offset A B C D SWIR Gain Setting A B C D SWIR Integration Time Setting the end frame of the file and the value of the time code that was received at the beginning frame The Time Code is constant over the range of frames included between the beginFrame and endFrame The values of the attribute show the beginning and ending frames e g 0 177 and the value of the time code from least to most significant number The time code values TC3 8 are decimal representations of hexadecimal numbers that reflect the UTC time TC5 8 are time codes values in seconds TC3 and 4 are sub seconds See previous attribute The Time Code values for corresponding SWIR and VNIR frames should be the same This attribute provides the VNIR offset settings which are fixed for the course of the EO 1 mission This attribute provides the SWIR offset settings which are fixed for the course of the EO 1 mission This attribute provides the SWIR gain settings which are fixed for the course of the EO 1 mission This attribute provides the SWIR integration time setting which is fixed for the course of the EO 1 mission The SDS attributes generated during the level 0 processing are carried over and appear with the level 1 files as well 3 5 Description of the Hyperion Level 1 Data Processing 3 5 1 Hyperion Level 1 Processing The Hyperion level 1 processing refers to the processing performed by TRW to produce the radiome
13. The Level code was updated to Level 1_A in response to Science Validation Team data needs and in response to knowledge gained during the Hyperion on orbit performance verification period The full list of changes incorporated in Level 1_A is provided in Section 3 5 1 Level 1 processing with the Level 1_A code began in July 2001 Timed with the release of the Level 1_A code was an update to the calibration file HypGain to HypGain_RevA The revision of the calibration file was an update to the method used to process the data used to generate the pre flight calibration file The revision improved the agreement in the VNIR SWIR spectral overlap region and extended the calibrated spectral range down to 400 nm Subsequent changes in echo correction parameters calibration file and bad pixel list are tracked using numbers in the data file appendages i e Ll A L1_Al Ll A2 and L1_A3 A readme file has been developed to track changes in the code as well as changes in the echo correction file ratio txt and the calibration file The current version of the readme file is Lilreadme_r3 xls which has been made available to the EO 1 Science Team for ready reference and has been attached to this document as Appendix 1 Updates of this file are prepared and sent to the EO 1 Mission Science Office for dissemination Revision B of this document includes a description of Revision B of the Level 1 processing code Level 1_B The Level 1_A c
14. all wavelengths while the absolute accuracy meets the requirements Table 1 Radiometric Performance Spectral Range Pre Flight On Orbit Absolute VNIR lt 6 consistent with Radiometry preflight end to end SWIR lt 6 consistent with preflight SNR 550 nm 150 192 650 nm 140 140 700 nm 140 140 1025 nm 90 65 1225 nm 110 96 1575 nm 89 64 2125 nm 40 38 Quantization All 12 bit 12 bit Table 2 presents the results of the spectral performance There are 220 unique spectral channels The baseline Level 1 processing calibrates 200 unique spectral channels and has an additional 4 channels of overlap Spectral bandwidths were measured precisely during TRW ground testing Direct on orbit measurements of these values were not attempted A technique using an atmospheric limb data collect was developed to verify the center wavelengths for the VNIR and SWIR spectral channels The number of spectral features in the SWIR portion of the data due to the combination of atmospheric lines and lines on the diffuse reflectance panel enabled verification of the center wavelength for the entire SWIR to 3 nm Although the number of available lines in the VNIR was more limited it was determined that the pre flight VNIR spectral calibration was still valid The results of the center wavelength verification were used to determine the cross track spectral error and the dispersion The on orbit measurement
15. data on a Digital Linear Tape DLT to TRW TRW performs Level 1 processing on requested DCEs Level 1 processing produces metadata files that can be used to track the processing steps as well as give an indication on the quality of the Level 1 data product TRW sends the Level 1 data and the metadata files to GSFC on a DLT This is described in more detail in chapter 3 GSFC handles the distribution of the data to the users Level 1 processing prior to July 2001 produced files with the extension L1 Level 1 processing began using Rev A of the Level 1 code at the beginning of July 2001 which produces image files with the extension Ll A Level 1 processing began using Rev B of the level 1 code on 15 November 2001 which produces image files with the extension L1_B Ancillary data in engineering units LO Science data Final product level 1 data metadata file attached Ship to GSFC Box Shape Key Hyperion processing Raw or GSFC processed data TRW function or product Figure 5 Hyperion Data Flow from GSFC to TRW 13 EO 1I Hyperion Science Data User s Guide HYP TO 01 077Rev Public Release L1_B Nominally it takes GSFC 3 5 days from receipt of the raw data to perform Level 0 processing and deliver the tapes to TRW TRW performs Level 1 processing and ships the final Level 1 data product with metadata files for select DCEs to GSFS within 3 days of receipt of the Level 0 data The length of time from
16. frame Averaged dark log file reports average value over all dark pixels for dark file used for dark subtraction L1 or Averaged pre and post image dark current log files L1_A Text file Calibration coefficient file Binary BSQ Echo removal log file one each for dark and image file Indicates file was corrected for echo Text file Smear correction log file one each for dark and image file Indicates file was corrected for smear Text file EO 1I Hyperion Science Data User s Guide HYP TO 01 077Rev Public Release L1_B Output Step in File Name Data File Description ref only Process and Format MD10 Step 6 EOIYYYYDDD VVVVSSSS _rX fix log Bad pixel replacement log file Indicates that file vvas corrected for known bad pixels The bad pixel locations band sample are contained in the log Text file Filename No of frames assessed Operational sensors VNIR SWIR Existence of streaking banding shading saturation focus problems line drops Cloud cover Averaged dark file values Text file MD12 Step 8 EOIYYYYDDD VVVVSSSS _rX LI hdr ENVI header file MD 13 Step 9 SpectralLO_revA 1 dat Full Spectral calibration contains center wavelength for each pixel 256x242 ascii MD 14 Step9 BandWidthL0_revA 1 dat Full Spectral calibration contains Gaussian full width half maximum for each pixel 256x242 ascii MD 15 Step 0 EOLYYYYDDD_VVVVSSSS _rX sat File containing location band sample frame of pixel valu
17. more detail the characteristics of the Hyperion data set The topics include the absolute calibration file as well as a description of the pixel to pixel variation the use of the spectral calibration file and the alignment of the VNIR and SWIR focal planes The subsequent chapter Chapter 5 discusses the end to end measurement error 4 1 Absolute Radiometric Calibration The absolute radiometric calibration and related topics is presented in the section The derivation and the verification of the calibration file are highlighted 4 1 1 Derivation of the Absolute Radiometric Calibration File The absolute radiometric calibration file generated on July 1 1999 was used as the baseline on orbit calibration file The calibration file used by the Level 1_A code is HypGain_revA and that used by Level 1 code is HypGain The difference between HypGain and HypGain_revA is the method used in generating the pre flight calibration file In both cases the solar calibration data collected on Day of Year 47 of year 2001 was used to remove pixel to pixel variations The solar calibration collection was also used to improve the VNIR SWIR overlap region for HypGain_revA The calibration file is used to process all Level 1 data files Updates to the calibration file if required will be accompanied with an explanation of the update and will be assigned a new revision designator 45 EO 1 Hyperion Science Data User s Guide HYP TO 01 077Rev Public Release Ll B
18. record of the intermediate processing steps The log files associated with the intermediate files retain the image filename followed by Jog The purpose of the log files is to indicate that the intermediate processing has occurred and in some cases indicate details of the processing that occurred Figures 12 and 13 shows which metadata files MD are produced during intermediate processing steps Selected metadata files are presented in the following section as examples In addition to the processing steps an image quality assessment is performed as a qualitative spot check on a single cube parsed from a requested DCE The results are reported in metadata file 11 MD11 The form is discussed in the next section 33 EO 1 Hyperion Science Data User s Guide HYP TO 01 077Rev Public Release Ll B Figure 12 Level 1 Data Processing Flow Diagram Receive LO data from GSFC Background removal Apply calibration gt Identify nearest gt Multiply smear echo corrected time wise avg dark LO image by calibration file to gt Subtract avg dark from obtain radiometrically corrected Smear correction gt Apply smear correction to SWIR data Both dark and image files 7 9 Level 1 data gt Log file generated MD9 image file dl i Log file generated MD5 gt Logfile generated MD2 gt Supply calibration file MD7 QA L1 image Echo removal FES gt Subset L1 DCEs gt Remove echo from smear Average dark files Fixstripes I
19. smear and echo artifact correction is required for all science applications The absolute calibration is not valid unless the smear and echo corrections have been performed and the level 0 data set has been properly processed The Level 0 output file also includes HDF attributes The HDF attributes include Scientific Data Set SDS attributes The subsequent sections discuss the Level 0 processing steps the ancillary HDF data files and the SDS attributes 3 4 1 Hyperion Level 0 Processing Level 0 processing of EO 1 Hyperion science data refers to the following set of tasks that are performed on data that has been downlinked from the spacecraft This set of tasks is performed by GSFC on all scenes that are collected A LO extension indicates that Level 0 processing has occurred The details are provided for those who require this level of detail For most users the most important thing to note is that Level 0 processing does not perform artifact correction on the SWIR data The corrections are required for absolute radiometric accuracy 1 Decode the data Perform Reed Solomon decoding on downlinked science data Extract the science data from the telemetry data and flag corrupted data 2 Separate the data into files Separate according to data type boundaries DCE image number boundaries and data stream boundaries Lunar calibration ground image solar calibrations are 25 EO 1 Hyperion Science Data User s Guide HYP TO 01 077Rev Public R
20. still be 242 spectral bands in each data file The number of along track pixels will mirror the frame number range will vary depending on file type e g dark image and may vary from image to image The number of along track pixels will be the same as the number of frames of data The frame rate is 223 4 frames sec A nominal ground image is 30 seconds long a dark file is 1 sec long and a lamp file is 3 seconds long The length of the ground image may vary The pixel order for all files is BIL Band Interleaved by Line Level 0 processing detects and flags pixels with missing data This attribute informs the user of how many pixels were found to have missing data although it does not identify which pixels were affected The sync time is the time from the spacecraft Time Code pulse to the Hyperion frame sync pulse with 32 usec resolution The values of the attribute show the beginning and ending frames e g 0 176 and the value of the sync time e g 39 See previous attribute The VNIR and SWIR receive their Time Codes and Sync Times separately The Time Code is the time broadcast by the spacecraft to Hyperion and is updated every second 2 msec The Time Code information included here shows the beginning frame of the file 29 EO 1 Hyperion Science Data User s Guide HYP TO 01 077Rev Public Release Ll B 6 TC7 TC8 SWIR Time Code beginFrame endFra me TC3 TC4 TC5 TC 6 TC7 TC8 VNIR Channel Offset A B C D
21. to help users not using ENVI get started The user is responsible for proper incorporation into processing codes The Level 1 data format is dependent on the version of the processing code so there will be two formats The original released Level 1 code data product with data suffix L1 was a 16 bit unsigned integer with units of radiance W m2 um sr times a factor of 100 The Level 1_A code with data suffix L1_A data format is 16 bit signed integer with units of radiance W m2 um sr times a factor of 40 for the VNIR bands 1 70 and a factor of 80 for the SWIR bands 71 242 The Level 1_B code has the same data format and units as Level 1_A but the VNIR and SWIR data have been spatially co registered For Matlab users the following lines are extremely useful The first grouping can be used as a template for a code in which one frame at a time can be read in The second grouping is useful if a subset of the full image is exported In this example 330 lines of one spectral band were subset to a file The data type would have to be changed for the updated level 1 products fid fopen fname r npf 242 256 count npf tmp count fread fid npf int1i6 img_tmp reshape tmp 256 242 imagesc img_tmp 21 EO 1 Hyperion Science Data User s Guide HYP TO 01 077Rev Public Release Ll B fid fopen fname r npf 330 256 count npf tmp count fread fid npf int16 img_bnd reshape tmp 256 33
22. who are using ENVI A brief discussion is provided in the last section for those who may use other software for processing and data analysis The user will typically receive the Level 0 files before the Level 1 processed file The Level 0 data set is described in chapter 3 2 1 Data Cube Structure A typical Hyperion image has the dimensions of 256x6925x242 The first number represents the number of pixels that span the field of view The span of the field of view defines the swath width One entire swath width of data is obtained for each frame The total number of frames is represented by the second dimension and defines the swath length The instantaneous field of view for each pixel and the frame rate 223 4 Hz define the dimensions of ground being imaged Each pixel location images approximately a 30 m by 30 m region of the ground The swath width for each focal plane is comprised of 256 pixel locations corresponding to 7 7 km There is a 1 pixel shift between the VNIR and SWIR cross track co registration resulting in a 255 VNIR SWIR coincident field of view locations For each pixel location 242 spectral channels of data are 14 EO 1 Hyperion Science Data User s Guide HYP TO 01 077Rev Public Release Ll B obtained Spectral channels from 1 70 are collected from the VNIR and channels 71 242 are collected from the SWIR Due to low signal for some channels and to reduce duplication in the VNIR SWIR overlap region some of these spectral ch
23. 0 imagesc img bnd For C programmers here is an outline in code indicating how to get access to data in a HDF file This is provided only in an attempt to assist users in accessing the data directly int sd_id SDstart hdf_file DFACC READ lt Open a HDF file to get an ID SDfileinfo sd_id amp nDataSets amp nFileAttrs lt General info nDataSets no of data sets nFileAttrs no of attributes prepare to get access to the data sets Retrieve data sets for int index 0 index lt nDataSets index int sds_id SDselect sd_id index lt Get an ID for the data set given by an integer of index SDgetinfo sds_id name amp rank dim amp data_type amp n_attrs lt Get general info of the data set use the info to set up proper parameters to get access the data set SDreaddata sds_id start NULL edge VOIDP buff lt Read the data set and store the data in buff data processing SDendaccess sds_id lt Close the data set further data processing SDend sd_id lt After the processing close the HDF file 3 HYPERION DATA COLLECTION FLOW AND PROCESSING This chapter describes the data collection event sequence the data flow and level 0 and level 1 processing steps The data collection event sequence and timeline is presented with a sample level 0 filename set The types of data collection events are described for users who rely on non ground
24. 3 1 dead 3 1 dead 4 1 dead 4 1 dead 5 1 dead 5 1 dead 6 1 dead 6 1 dead 7 1 dead 7 1 dead 8 1 dead 8 1 dead 9 1 dead 9 1 dead 10 1 dead 10 1 dead 11 1 dead 11 1 dead 12 1 dead 12 1 dead Status 13 1 dead 13 1 dead dead pixel value 0 14 1 dead 14 1 dead flat non responsive same value entire 15 1 dead 15 1 dead scene 16 1 dead 16 1 dead new bad pixels 17 1 dead 17 1 dead 18 1 dead 18 1 dead 19 1 dead 19 1 dead 20 1 dead 20 1 dead 21 1 dead 21 1 dead 22 1 dead 22 1 dead 23 1 dead 23 1 dead 24 1 dead 24 1 dead 25 1 dead 25 1 dead 26 1 dead 26 1 dead EO 1I Hyperion Science Data User s Guide HYP TO 01 077Rev Public Release L1_B bad pixel file history 59 EO 1 Hyperion Science Data User s Guide HYP TO 01 077Rev Public Release L1_B Gain File Modifications Gain file history A hypgain HypGain revA dat HypGain revB dat HypGain revB dat Calibrated VNIR bands nm 9 57 437 925 5 57 400 925 8 57 427 925 VNIR SWIR Calibrated SWIR bands nm 75 225 890 2400 75 225 890 2400 77 224 912 2395 5 75 Total number of Calibrated bands 200 204 198 6 76 Number of Unique calibrated bands 196 200 196 7 225 Number of VNIR SWIR overlap bands 4 4 2 60
25. 6 pixels x 242 bands x 206 frames Data type 16 bit unsigned integer Byte order big Pixel order BIL Output file File name drb12 archive levell ground E012000147_01D101D0_rl avg Type Generic Dimensions 256 pixels x 242 bands x 1 frames Data type 16 bit unsigned integer Byte order big Pixel order BIL Source drb12 archive levell ground E012000147_01D101D0_r1l echo Summary Average 523 625 KKKKKKKKKKKKKKKKKKKK Figure 15 Example MD5 Average Log MDSA and MDSB have the same format 40 EO 1 Hyperion Science Data User s Guide HYP TO 01 077Rev Public Release L1_B The Echo Removal Log MD8 has the same general format as Average Log MDS but it reports that the echo correction process has been completed At the end of this log file the specific echo correction file used in the correction process is identified ratio txt along with the number of the band at which echo correction began The VNIR focal plane covers spectral bands 1 70 so it is expected that echo correction which is required only in the SWIR would start at band 71 MD8 EO12000147_01CA01C9_r1 echo log KKKKK hypecho KKKKK HIP 1 1 Mon Jun 5 13 06 05 2000 Arguments source file drb12 archive levell ground E0O12000147_01CA01C9_r1 smear source type hyperion output file drb12 archive levell ground E012000147_01CA01C9_r1 echo output type hyperion start band 71 ratio file dra2 calfiles ratio txt verbose
26. ILS oosi sess cscs soceccsicessssnccctcsccsscssccvecececsssentescessdecesssucteccasceSevsbnctascosecscsoceoseosseests 45 4 1 ABSOLUTE RADIOMETRIC CALIBRATION c ccccccssssssssssssssssssssssssssssscssscssscsssesessscsesesssssesssesssesssesesesessseseseseseees 45 4 2 RADIOMETRIC CALIBRATION ADVANCED TOPICS 2222222222222ununuaananananananananananananananenaneninenenininninininnn 46 4 3 SPECTRAL CALIBRATION sessseererercrertrtrtrtrtrtrtrtr trer trette tetetete tr rtr r e e e e EEn E EEE EEE E EEE E EEE EEE EEE EEE E EEE EEEE Eeee 47 44 VNIR SVVIR SPATIAL ALIGNMENT ADVANCED TOPICS LEVEL 1 AND LEVEL 1_A DATA PRODUCTS ONLY 51 END TO END MEASUREMENT ACCURACY sccscsscssssssscsscccccsssssceccccccscssssceeccccseesssccecescesscsscececescsseces 52 EO 1 Hyperion Science Data User s Guide HYP TO 01 077Rev Public Release Ll B List of Figures Figure 1 Schematic of the satellite constellation This figure shows the overlap and dimensions of the groundtracks for Landsat 7 and the three EO 1 instruments Hyperion Advanced Landsat Imager and the Atmospheric Corrector Note that EO 1 follows one minute behind Landsat l the AMOS ALOE OIE eieeei esmenat one ea esti es ene 8 Figure 2 Drawing of Hyperion Instrument and Electronics ee eee eeeeeseeceseceeeeeeeeeceaecneeeeeeeeaeees 9 Figure 3 Photo of the Hyperion Sensor Assembly HSA 0 eeceesceesseceseeceseeeseeeeneecsaecnseeeseeesaeees 9 Figure 4 Photo of Hyp
27. LO_revA there is a variation of the center wavelength across the field of view that should be taken into account in certain applications Figures 21 and 22 show the variation across the field of view for selected VNIR and SWIR wavelengths The wavelengths selected span the calibrated range of wavelengths and should give the user a sense for the variation for the entire focal plane The variation of the spectral smile is smaller for the SWIR than it is for the VNIR The spectral calibration is applicable for the life of the mission 47 EO 1I Hyperion Science Data User s Guide HYP TO 01 077Rev Public Release L1_B Center Wiwsengin inm boo 400 50 100 150 200 Spectral Channel Figure 20 Image of the Center Wavelength Calibration File 48 EO 1I Hyperion Science Data User s Guide HYP TO 01 077Rev Public Release L1_B 50 100 120 200 Spectral Channel Figure 21 Image of the Full Width Half Maximum Calibration File 49 EO 1I Hyperion Science Data User s Guide HYP TO 01 077Rev Public Release L1_B VNIR Spectral Variation Across the Field of View VNIR Band 10 447 892 VNIR Band 30 651 278 VNIR Band 55 905 511 Wavelength Relative to FOV 128 nm in 0 50 100 150 200 250 Pixel Field of View Figure 22 Variation of the VNIR center wavelength across the field of view SWIR Spectral Variation Across the Field of View e SWIR Band 75 892 35
28. NIR and 10 counts in the SWIR The effect of 46 EO 1 Hyperion Science Data User s Guide HYP TO 01 077Rev Public Release Ll B scatter has been estimated and is included in the error budget It should be noted that the when the cover is closed there is no measurable light leakage into the instrument 4 2 4 Saturation VNIR saturation occurs on rare occasions on the top of clouds over a limited spectral extent SWIR saturation has been limited to hot spots resulting from gas flares or volcanoes The revised level 1 processing Level 1_A flags saturated pixels It should be noted that the SWIR echo and smear algorithms cannot remove the artifact when the source pixel is saturated 4 3 Spectral Calibration The derivation and verification of the spectral calibration file is discussed 4 3 1 Spectral Calibration files The spectral calibration for Hyperion is described by a Gaussian bandpass and is defined by a center wavelength and full width half maximum defined by SpectralLO_RevA 1 and BandwidthLO_RevA 1 respectively Both files are provided with the level 1 processed data set The size of each file is 256x242 since the spectral calibration is extended for every pixel in the VNIR and SWIR focal plane The bandwidth of each pixel is approximately 10 nm The spectral calibration supplied by SpectralLO_RevA 1 and BandvvidthLO RevA l is presented as images below Figure 19 and 20 respectively Although not evident in the image of the Spectral
29. NTACT LUIS Ti iii dte i BETS 14 HYPERION DATA CUBE QUICK START aiacseccsdscsascesccscssedsccdvec sdenteccdssseesdesvesceducsavossdecsdsesseccecesueseecesecesestess 14 2A DATA CUBE S TRUCTURE quocient Rotecna apace tite 14 2 2 SUGGESTED BAND COMBINATIONS FOR IMAGES ccccccccccccccccccccscsecesesesesesesesesesesesesesesesesesesesesesesseesesesessesseuess 15 2 3 INTRODUCTION TO THE HYPERION SPECTRA uu cccccccccccccccccccscccecsecesescceseseseseseseseseseseseseseseseeeseseseseseeeesseseeseseseseess 18 2 4 AUXILIARY PLATFORM NOTES OPTIONAL ccccccccccccccccccccscscsecesesesesesescsesesesesesesesesesessseeeseseseseeessssseseseseeseeeeeess 21 HYPERION DATA COLLECTION FLOW AND PROCESSING ccssssssssssssccsssscccssseccscssscecessssccessnee 22 3 1 HYPERION DATA COLLECTION EVENT SEQUENCE cccccsessscecececeesensscecececseseseeecececeesesseaeeeceeeeseesaeeeeeceesensaaeas 23 3 2 TYPES OF HYPERION DATA COLLECTION EVENTS ccccccscececececececececececececececececececececececeescececececseeeececececececesess 23 3 3 DATA FLOW FROM THE SPACECRAFT TO THE USER ccccccesecesesececececececececececececececececececececececececeeseecececeeeceeesecs 24 3 4 DESCRIPTION OF THE HYPERION LEVEL 0 DATA PROCESSING ccccccccceseceeesecececececececececscscseseeeececeeeeseseseeeeeess 25 3 5 DESCRIPTION OF THE HYPERION LEVEL 1 DATA PROCESSING ccccecccececeeececececececececsescseececscseseseseseseeeeeeeeeess 30 HYPERION DATA CUBE DETA
30. TRW Space Defense One Space Park R VV amp Information Systems Redondo Beach CA 90278 CAGE No 11982 TITLE EO 1 Hyperion Science Data User s Guide Level 1_B DATE November 2001 NO HYP TO 01 077 REV Pub Release L1_B Superseding Pub Rel Ll A 8 01 PREPARED BY Pamela Barry Date Hyperion Performance Analysis and Team Lead APPROVAL SIGNATURES Carol Segal DATE Steve Carman DATE TRW Hyperion Deputy Project Manager TRW Hyperion Project Manager Mission Operations and Planning DATE DATE ORIGINAL PDMO RELEASE DATE EO 1 Hyperion Science Data User s Guide HYP TO 01 077Rev Public Release Ll B 1 5 TABLE OF CONTENTS INTRODUCTION nicsavsssosssdsscccnsonsesessonsoecedsesovsuovsceedcseodesscsscsacseevssecdetaesoesesessecbaasodesseseuconteseessecceesseasesessvesvecssoeseseeses 6 1 1 DOCUMENTS COPE asil ated a whee Bega a Er TO se i 6 1 2 EO 1 MISSION OVERVIEW cccccccececscecececececececececececececececececececececececscecececececececececececeeeceseceeeceeeeseecececececeeeceeecens 7 1 3 HYPERION INSTRUMENT OVERVIEW cccccccecececececececececececececececececececececececececececececececececececececececececececececececececevees 8 1 4 SUMMARY OF HYPERION PERFORMANCE CHARACTERISTICS cccecesecececececececececececececececsceceeeceescseseseeeseseseeeess 10 1 5 OVERVIEW OF HYPERION DATA COLLECTION AND DATA FLOW ccccccccececesecececececececececececesecececseeeseeceeseeeeess 12 1 6 HYPERION CO
31. a S a E ai 37 Hyperion End to End Measurement Error sseseeeeseseeseesessesererressersresreeseesersersseeseesersseesee 52 EO 1 Hyperion Science Data User s Guide HYP TO 01 077Rev Public Release Ll B 1 INTRODUCTION Hyperion is a hyperspectral instrument on the Earth Observing 1 EO 1 spacecraft that was launched from Vandenberg Air Force Base on November 21 2000 EO 1 is part of NASA s New Millennium Program which is an initiative to demonstrate advanced technologies for dramatically reducing the cost and improving the quality of instruments and spacecraft for future space missions Under this program missions are intended to validate new technologies in flight and to provide useful scientific data to the user community The primary demonstrations are oriented towards remote sensing technologies and spacecraft technologies that will be used in defining future Landsat type missions The instrument payloads on the spacecraft are Hyperion ALI Advanced Land Imager and AC atmospheric corrector The first three months of the mission life were focused on instrument activation and performance verification This document EO 1 Hyperion Science Data User s Guide introduces the user to the Hyperion data set details the data processing steps and highlights performance characteristics from the standpoint of the user Revision A of this document includes the description of the release of Revision A of the Level 1 processing code Level 1_A
32. al of the Hyperion is to provide a science grade hyperspectral instrument with quality calibration based on existing critical designs and existing selected hardware Hyperion also supports the evaluation of ALI LAC and comparison with LANDSAT ETM The EO 1 has a sun synchronous orbit with an altitude of 705 km and a 10 01 AM descending node The orbit inclination is 98 2 degree the orbital period is 98 9 minutes and the EO 1 equatorial crossing time is one minute behind Landsat 7 The velocity of the EO 1 nadir point is 6 74 km sec Figure depicts the formation flying capability of the EO 1 spacecraft Also depicted is the overlay of the swath width for the different instruments EO 1 Hyperion Science Data User s Guide HYP TO 01 077Rev Public Release Ll B wae _ Multispectral Imager 36 km 30 m Landsat Figure 1 Schematic of the satellite constellation This figure shows the overlap and dimensions of the groundtracks for Landsat 7 and the three EO 1 instruments Hyperion Advanced Landsat Imager and the Atmospheric Corrector Note that EO 1 follows one minute behind Landsat in the Landsat orbit 1 3 Hyperion Instrument Overview The Hyperion instrument provides radiometrically calibrated spectral data The purpose of the data is to support evaluation of hyperspectral technology for Earth observing missions Hyperion is a pushbroom imaging spectrometer Each ground image contains data for a 7 65 km wide cross track by 185 k
33. annels are not calibrated The uncalibrated channels are set to zero The zero ed channels are not removed from the file so the final data set is the same size as the initial data set There are three versions of the Hyperion Level 1 data product suffixes L1 L1_A and L1_ B so the header and the data file extension should be reviewed to determine which data product is being analyzed Original level 1 The data is an unsigned integer The data is presented as calibrated radiance W m2 sr um times a factor of 100 for both the VNIR and the SWIR The calibrated data file has the extension L1 Revision A The data is a signed integer A scaling factor has been applied to the calibrated radiance VV m2 sr um A factor of 40 was applied to spectral bands 1 70 and a factor of 80 was applied to spectral bands 71 242 To obtain data in units of mW cm2 sr um the data should be multiplied by 107 The extension to the calibrated data file is L1_A The header file will also indicate the version of the processing code as well as the factors used for the VNIR and SWIR bands Revision B The SWIR and VNIR components of the data have been spatially co registered in the cross track and along track dimensions An additional metadata file a text file with the extension aln log indicates the source file and the output file names for the final co registered data product L1_B 2 2 Suggested Band Combinations for Images Using ENVI the following i
34. as been studied and is consistent with pre flight characterization The remaining categories are not topically present An X in the box under SWIR indicates that the quality being assessed e g streakiness unevenness across the FOV was observed somewhere within the first 660 frames of the SWIR image The average dark file value is reported from MDS MD11 EO12000999_01CD01CC_r1 L1 ga File name E012000147 O01CDO1CC r1 L1 Frames assessed 1 660 Source tape ID E01189 Sensors operational X VNIR X SWIR 1 0 Image Quality Check Comments Streaking x none Banding Shading across FOV Focus Saturation Linedrops I 2 0 Radiometric Calibration 2 1 Averaged dark file value 523 6 3 0 Additional Comments Figure 14 MD 11 Quality Assessment Form 39 EO 1 Hyperion Science Data User s Guide HYP TO 01 077Rev Public Release L1_B 3 5 3 Description of Sample Metadata Files This section describes the contents of sample metadata files Once the user is familiar with the structure of the metadata file the user should be able to gain information from all of the metadata files The first three lines of a log file identify the Level 1 processing command from which the log file was generated the Hyperspectral Image Processing HIP software version number and date and time file was processed In this case the command used to create the file is cubeavg version 1 1 of HIP was used and the level 1 pr
35. ata using the Level 1_A code L1_B processed data using the Level 1_B code MD Metadata file number Output Step in File Name ref only Process L1 Completion EO1YYYYDDD VVVVSSSS IX LI of Step 6 Ll A Completion EO1YYYYDDD VVVVSSSS IX LI A of Step 6 Ll B Completion EO1YYYYDDD_VVVVSSSS_rX L1_B of Step 7 MDI Step 9 HYP PREFLIGHT RX txt MD2 Step 5 EOIYYYYDDD VVVVSSSS _rX cal log MD3 Step 3 EOLYYYYDDD_V V V V S S S S _rX avg L1 MD3A MD3B L1_A MD5 Step 3 EOIYYYYDDD V V V V SOSS S rX avg LD log MDSA MDSB LI A MD7 Step 5 HypGain_revA bin MD8 Step 2 EOIYYYYDDD VVVVSSSS _rX echo log EOLYYYYDDD_V V V V S S S S _rX echo log MD9 Step 1 EOIYYYYDDD VVVVSSSS rX smear log EOLYYYYDDD_V V V V S S S S _rX smear log 37 Data File Description and Format Fully processed Level 1 DCE Data Collection Event HDF band interleaved by line BIL order Fully processed Level 1 DCE Data Collection Event HDF band interleaved by line BIL order Fully processed Level 1 DCE Data Collection Event HDF band interleaved by line BIL order Summary of pre flight instrument characterization Text file MS Word Level 1 calibration log file Indicates file was calibrated Text file Averaged dark current file L1 ave over no of dark frames acquired or averaged pre and post image dark current files L1_A 256 x 242 pixels Binary BSQ equivalent to BIL for a single
36. by 40 0 and dividing the data in bands 71 242 by 80 0 18 EO 1 Hyperion Science Data User s Guide HYP TO 01 077Rev Public Release L1_B Vegetation ES Roof Top a Water Yellow Flat c d Figure 8 Locations of San Francisco Salt Ponds 1 17 01 used for showing spectral examples 19 EO 1 Hyperion Science Data User s Guide San Francisco January 17 2001 Roof Top HYP TO 01 077Rev Public Release Ll B HypGain_revA 160 140 Reflection of Solar Irradiance of a roof top a relfective surface Ra Some atmospheric features are dia 120 indicated nc e 100 water W Oxygen absorption Line band m 80 2 u m water 60 absorption sr band water water VNIR SWIR absorption F a Spectral band absorption utilized Overlap band co spectral 20 Region i channels non utilized 1 spectral 0 rte he 350 1350 1850 2350 Wavelength nm Figure 9 Example of a Hyperion Spectrum 20 EO 1I Hyperion Science Data User s Guide HYP TO 01 077Rev Public Release L1_B San Francisco January 17 2001 Sample Spectrum HypGain_revA Roof Top a Water c Yellow Flat d Vegatation f 160 140 120 100 Radiance W m2 um sr 1200 1400 1600 1800 2000 2200 2400 Wavelength nm Figure 10 Additional examples of Hyperion Spectrum 2 4 Auxiliary Platform Notes optional This section is provided
37. ction vvvv and ssss are the hex representations of the two byte VNIR and SWIR file IDs This is used to identify the instrument the focal plane the image number and the type of collect The decoding of the hex representation is a complex process and is out of scope for this document The information contained in the hex representations is already decoded for the user and is contained in the SDS attributes provide below r1 indicates this is the first run of this data set through the data processing software ggg is an identifier indicating from which ground station the data were received XX represents the number of downlink attempts for ground station ggg 3 4 3 Ancillary HDF Data Files Included on a DLT with the level 0 Hyperion data are a set of HDF files referred to as ancillary data as introduced above The ancillary data are a subset of spacecraft and instrument telemetry recorded during the related DCE The list of Hyperion mnemonics included in the ancillary data is presented in Table 6 along with a brief description and typical values The Hyperion HDF file names follow the format yyyy_ddd_hyp hdf indicating the year yyyy and Julian date ddd of the data which has been recorded for all the DCEs taken on that date The telemetry values are recorded in engineering units where applicable The values for the mnemonics are reported in the ancillary data along with the corresponding Mission Elapsed Time MET in seconds
38. ction 3 4 2 Table 4 Basic Data Collection Event Timeline Event Mm ss Product Example LO name relative to Ex LakeFromeDay005 Year2001 scene start Standby 10 31 Dark collect 00 31 Dark 1 EO12001005_122A1229_r1_PF1_01 L0 start Pre Image Dark Dark collect 00 30 stop Cover open 00 28 Image start 00 03 Scene EO12001005_122D122C_r1_PF1_01 L0 Variable length 00 00 image Image stop 00 03 Cover closed 00 11 Dark collect 00 29 Dark 2 EO12001005 122F122E r1_PF1_01 L0 start Post Image Dark Dark collect 00 30 stop Lamp on 00 32 Lamp collect 03 32 Lamp Collect EO12001005 12371236 r1 PFI O1 LO start Lamp collect 03 35 stop Lamp off 03 37 Dark collect 03 57 Dark 3 EO12001005_12341233_r1_PF1_01 L0 start Post Lamp Dark Dark collect 03 58 stop Idle 04 00 3 2 Types of Hyperion Data Collection Events Although there is one standard DCE sequence there are subtleties in the details of the collection Table 5 For example the standard ground image collect is 30 seconds with the center 24 seconds being the primary region of interest The solar calibration DCE is nominally 16 seconds 23 EO 1 Hyperion Science Data User s Guide HYP TO 01 077Rev Public Release Ll B In addition to the length of each image there are subtleties in the spacecraft pointing and motion during the collect For the ground collect the spacecraft pitch yaw is commanded
39. d to be 30 67 meters in the cross track direction and 30 56 meters in the along track direction for each pixel The swath width is spanned by 256 field of view locations and there are 242 spectral channels for each field of view location This results in 61 952 different pixels The complexity is that 51 EO 1 Hyperion Science Data User s Guide HYP TO 01 077Rev Public Release Ll B each pixel in the VNIR and SWIR view a slightly different portion of the ground in a single instance 4 4 2 VNIR and SWIR Spatial Co registration of Spectral Channels The spatial co registration is a measure of an object s position in the FOV as a function of the spectrometer wavelength The spatial co registration for the VNIR is better than 0 25 pixels The spatial co registration for the SWIR is better than 0 28 pixels in the spectral direction This characterization treats each focal plane as separate units 4 4 3 Co registration between the VNIR and SWIR It is understood that the user may have the desire to look at the entire spectrum for a single spatial location In order to do this the spatial co registration between the VNIR and the SWIR becomes important There is a difference between the VNIR to SWIR co registration for the cross track and in track direction For the cross track direction there is a one pixel offset between the VNIR and SWIR So pixels 1 255 in the VNIR correspond to field of view pixels 2 256 in the SWIR The difference in the in t
40. d user Also presented is the saturation limit of the instrument and the rationale behind the VNIR and SWIR multiplication factors for level 1 processing 4 2 1 Proper Operation Temperature of the SWIR The SWIR FPE temperature varies from DCE to DCE within the temperature range mentioned above The responsivity of the SWIR is weakly sensitive to SWIR FPE temperature within this range The variation due to temperature sensitivity is included in the repeatability estimate and applies to temperature variations in the range of 153 5 C 1 C 4 2 2 Pixel to Pixel Variations The on orbit calibration file was field flattened based on the solar calibration event that took place on Day of year 47 of 2001 The pixel to pixel variations which are constant throughout a DCE may vary slightly from DCE to DCE All of the scenes are being field flattened based on the Day of year 047 2001 solar calibration 4 2 3 Optical Scatter During the processing of the raw VNIR solar calibration data after dark field removal the existence of an additional offset in the dark field appeared at wavelengths in the very blue lt 400 nm and the infrared lt 950 nm It was noticed because there was obviously no real spectral signal in the dark corrected counts The effect was not noticed at spectral pixels in the mid wavelength range where the solar response is strong The size of the scatter for a solar calibration data collection is on the order of 70 counts in the V
41. data contains this parameter This is discussed further in chapter 3 The absolute calibration for the SWIR is only applicable at the operational temperature The SWIR has negligible response when too warm If the SWIR image does not contain features consistent with the VNIR image then the SWIR was quite possibly not at the operational temperature However it may not be clearly evident from the image if the SWIR is 17 EO 1 Hyperion Science Data User s Guide HYP TO 01 077Rev Public Release Ll B slightly off operational temperature The SWIR FPE temperature is maintained by a cryocooler Due to operational constraints the cooler does not run continuously A table on the EO 1 web site also contains the on and off times of the cryocooler This may be used as a quick look tool 2 3 Introduction to the Hyperion Spectra Sample spectra from the San Francisco scene were extracted Figure 8 indicates which region of the scene each spectrum represents Figure 9 contains the spectrum for sample A a rooftop The spectrum is influenced by the solar radiance spectral profile which resembles a black body of 6000K temperature with Fraunhoffer lines modified by the reflection from the surface and atmospheric effects Some atmospheric features are annotated in the Figure 9 The oxygen line CO and water absorption features are the most prominent Also indicated in the Figure 9 is the overlap region of the VNIR and SWIR as well as the spectral channels that ar
42. e Data User s Guide HYP TO 01 077Rev Public Release L1_B MD10 EO12000147_01CD01CC_r1 fix log KKKKK cubefix KKKKK HIP 1 1 Mon Jun 5 13 08 29 2000 Arguments source file drb12 archive levell ground E0O12000147_01CD01CC_r1 cal source type hyperion output file drb12 archive levell ground E012000147_01CD01CC_r1 L1 output type hyperion bad file dra2 calfiles badpix txt verbose Source file File name drb12 archive levell ground E012000147_01CD01CC_r1 cal Type Hyperion Dimensions 256 pixels x 242 bands x 660 frames Data type 16 bit unsigned integer Byte order big Pixel order BIL Output file File name drb12 archive levell ground E012000147_01CD01CC_r1 L1 Type Hyperion Dimensions 256 pixels x 242 bands x 660 frames Data type 16 bit unsigned integer Byte order big Pixel order BIL Source drb12 archive levell ground E012000147_01CD01CC_r1l cal Bad pixels file File name dra2 calfiles badpix txt Bad pixels band sample 2 3 4 5 6 K 8 9 PRPRPPRPRPER OCI I BCN FP OO EO 1 Hyperion Science Data User s Guide HYP TO 01 077Rev Public Release Ll B OoOWMAANIAOPWNHE NONNNNNNN N NONNNNNNNNNNNNNNNNNNNNNNDN N N pis 92 and 94 94 and 96 r 0 iy 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 1 2 4 OOO 92 and 94 91 and 93 138 50 of 137 and 139 256 100 of 255 23 50 of 22 a
43. e not currently used Figure 10 displays additional spectral profile examples Subtleties of the spectra and working with the spectra are described in chapter 4 For example while these spectra are illustrative of general trends detailed analysis must include consideration of spectral smile and spatial co registration The wavelength included in the header file was used with the z profile spectrum ENVI tool to create these plots The spectrum was saved to an ascii file and then loaded into Excel This process assigns the center wavelength included in the header to the spectrum This wavelength is only accurate for the field of view pixel 128 Applications that want to take advantage of the spectral calibration need to use the SpectralLO_revA dat file that is provided on the level 1 tape This file contains the center wavelength for every single pixel This is discussed more in a later chapter The spectral profiles presented in this section are a combination of the VNIR portion for one field of view location and the SWIR portion for the corresponding field of view location For applications in which the scene is highly variable spatially and when working with either L1 or L1_A data attention must be paid to the VNIR SWIR co registration to make sure the combined spectrum truly represents the location of interest L1_B data is spatially co registered The Ll A data was also returned to absolute radiance scale by dividing bands 1 70 in the VNIR
44. ed 0 222 integer Dimensions Number of Cross 1 16 bit integer 256 Track Pixels Number of Bands 1 32 bit integer 242 Number of Frames 1 32 bit integer 223 Pixel Order 8 Bit character BIL Number of Missing Pixels 1 32 bit unsigned 0 integer VNIR Sync Time 3N 32 bit integer 0 176 39 beginFrame endFrame syncTime 177 222 121 SWIR Sync Time 3N 32 bit integer 0 176 39 beginFrame endFrame syncTime 177 222 121 VNIR Time Code beginFrame 8M 32 bit integer 0 177 143 0 224 endFrame TC3 TC4 TC5 TC6 204 214 5 TC7 TC8 178 222 143 0 225 204 214 5 SWIR Time Code beginFrame 8M 32 bit integer 0 177 143 0 224 endFrame TC3 TC4 TC5 TC6 204 214 5 TC7 TC8 178 222 143 0 225 204 214 5 VNIR Channel Offset A B C D 4 8 bit unsigned 8 8 8 8 integer SWIR Channel Offset A B C D 4 8 bit unsigned 97 97 104 102 integer SWIR Gain Setting A B C D 4 8 bit unsigned 1 1 1 1 integer SWIR Integration Time Setting 1 8 bit unsigned 125 integer Image attribute Level 0 File Generated By Byte Order This attribute describes what type of data is contained in the data file e g pre image dark cal image post image dark cal lamp cal post lamp dark cal This attribute defines which version of the level 0 processing code was used to process the data file e g HLZP version 1 0 0 This attribute indicates that the file byte order is big endian rather than little endian Different hardware software combina
45. eduled consistent with existing priorities The spacecraft would collect the scene The Hyperion science data and ancillary data obtained during the DCE would be stored on the WARP The Hyperion science data includes the five data files described above and the ancillary data refers to the instrument telemetry obtained during the DCE The WARP also stores science data from the other instruments as well as ancillary data from the other instruments and select subsystems The science data and ancillary data are downlinked to one of several ground stations using an X band downlink All of this data is recorded on Ampex tape and sent to GSFC by the ground stations GSFC performs Level 0 processing on the data Description of this processing is provided in the next section GSFC sends the Level 0 data and ancillary data on a DLT to TRW and indicates which scenes should be processed to Level 1 GSFC also sends the Level 0 data to the user before the Level 1 data has been produced TRW performs Level 1 processing on requested DCEs Level processing produces metadata files which can be used to track the processing steps as well as give an indication on the quality of 24 EO 1 Hyperion Science Data User s Guide HYP TO 01 077Rev Public Release Ll B the Level 1 data product TRW sends the Level 1 data and the metadata files to GSFC on a DLT GSFC distributes data to the user Nominally it takes GSFC 3 5 days from receipt of the raw data to perform Lev
46. el 0 processing and deliver the tapes to TRW TRW performs Level 1 processing and ships the final Level 1 data product with metadata files for select DCEs to GSFC within 3 days of receipt of the Level 0 data GSFC will then forward the Level 1 data to the people who request the scene It is possible for the user to receive a tape from GSFC that contains the Level 0 Hyperion data and then later receive a tape that contains the Hyperion Level 1 data The Level 0 should not be used for science applications because artifact corrections have not been performed The following is a highlight of the process of the Hyperion data flow described above 1 Scene request submitted 2 Scene scheduled 3 Scene collected and stored on WARP 4 Scene downloaded to a ground station 5 Ground Station sends data to GSFC 6 GSFC performs Level 0 processing 7 GSFC sends Level 0 data to TRW on a DLT and sends Level 0 data to the requestor 8 TRW performs Level 1 processing on requested scenes 9 TRW sends Level 1 data to GSFC on a DLT 10 GSFC sends Level 1 data sets to the requestor 3 4 Description of the Hyperion Level 0 Data Processing Level 0 processing of EO 1 Hyperion science data refers to the following set of tasks that are performed on data that has been downlinked from the spacecraft The processing turns the downlinked data into a set of Level 0 files with prescribed file names The Level 0 processing does not include artifact correction The SWIR
47. elease Ll B examples of data type boundaries Pre image dark image post image dark are examples of data stream boundaries 3 Perform checks to verify data integrity and instrument performance If a check fails then the program terminates with a descriptive error message 4 Combine the VNIR and SWIR data sets Arrange VNIR SWIR pixel order Confirm that the VNIR and SWIR file pair up properly Verify that the two filenames have the same year and day stamp represent the same DCE image number same event type and same data type i e pre image dark cal image post image dark cal or lamp cal Concatenate the VNIR and SWIR science data together 5 Create HDF Send the science data into a formatted output file HDF format Append HDF attributes to the output file These attributes which are described below in detail list various properties about the data 3 4 2 Hyperion Level 0 Output File Naming Convention The Level 0 processing output for Hyperion consists of a set of five files with the LO extension for each DCE Data in this output file is referred to as Hyperion Level 0 data The output filename is in this format EOlyyyyddd_vvvvssss_rl_ggg XX_ LO for example EO12001005_122A1229_r1_PF1_01 L0 is the pre image dark taken January 5 2001 The yyyyddd date is the UTC or Julian date of the beginning of the DCE collection where yyyy is the year and ddd is the day of year with January 1 corresponding to 001 The next se
48. erion as mounted on the EO 1 spacecraft 0 0 ee eee eeseeeseeeeeeeeeeenseeeeeeenees 10 Figure 5 Hyperion Data Flow from GSFC to TRW 0 eee eeeceseceseceseeesseeceaeceseesseeesaeecnaeenseensees 13 Figure 6 Example Gray Scale display of Hyperion data San Francisco Salt Ponds 1 17 01 16 Figure 7 Example of RGB viewing of Hyperion Data San Francisco Salt Ponds 1 17 01 17 Figure 8 Locations of San Francisco Salt Ponds 1 17 01 used for showing spectral examples 19 Figure 9 Example of a Hyperion Spectr unces ches sadccssesysons tesa tvesd e dy teas obvecsadee tance teqnnes tase kecenaeaaeen sare 20 Figure 10 Additional examples of Hyperion Spectrum eeccceesceeesececeeececeeeeeceeeeeeseeeeneeeenaees 21 Figure 11 VNIR SWIR Spatial Co registration ccsscesccsesscsessnssesnncsssesesensccsenacessnaccesseccesneess 32 Figure 12 Level 1 Data Processing Flow Diagram sessssssesssessessseessesessstesseesseesseessseeesseesseesse 34 Figure 13 Level 1_A Data Processing Flow Diagram e ssssssssssssesssesssseeesstessresseesseresseessseessesse 35 Figure 14 MD 11 Quality Assessment Form sessesseseesessesesesressessrssresseserestessessresrenseesresrresreseee 39 Figure 15 Example MD5 Average Log MDSA and MDSB have the same format 40 Figure 16 Example MD 8 Echo Removal Log esssssssesesseseesessrssresseseresressessrerrenseesrerreesreseese 41 Figure 17 Example MD 9
49. es gt 4095 for echo smear and image file Text file MD 16 Step 9 O0indexH YXXXX Listing of Level 1 DLT tape contents for tape number HY XXX X Text file MD 17 Step 7 EOLYYYYDDD_VVVVSSSS_r1_XXX aln l o VNIR SVVIR co registration log file es Indicates L_1B file was co registered Text file 3 5 2 Description of Quality Assessment Form MD11 is a Quality Assessment form shown in figure 14 which is filled out manually and after inspection of a portion of the DCE using ENVI The form indicates the file name of the scene being examined and which specific frames were used in the assessment The beginning of the image is frame 1 and there are 220 frames per second of image The form also indicates the number of the DLT tape on which the Level 0 data was received from GSFC and which sensors were operational for the image For virtually all DCEs both sensors are operational Operational however does not necessarily mean that the SWIR was at the proper operational temperature The average dark value can be used as a rough reference the cooler on off schedule can also be used as a rough reference The true reference is the SWIR FPE temperature that is contained in the HDF Level 0 file described in section 3 4 3 The mnemonic is YSWIRFPET and the Hyperion SWIR is at the proper operational temperature when the SWIR FPE temperature is 153 5 1C 38 EO 1 Hyperion Science Data User s Guide HYP TO 01 077Rev Public Release L1_B The qual
50. esponsive same value for entire scene pixels to bad pixel list badpix3 pepee Lee neighboring pixels contain a positive and a negative value the dead pixel was incorrectly replaced with values close to 32768 To correct for this error the method used by cubefix to average adjacent pixels was modified This mitigated the occurrence of data overflow when taking the average of a positive and negative number Saturated Sena a A modified version of cubesat repairs errors in the translation of saturated pixel locations to band sample and frame numbers Es modified version accurately reports band and frame number but the sample number is off by plus 1 L1 A2 Gain file JHypgain RevB dat removes zeros bands 5 6 7 75 76 and 225 from the L1 A2 data product LE Min Max In the L1_A and L1 A1 data products the maximum value Cal max of a calibrated pixel was set to 32768 which is not a valid signed integer The range of 16 bit signed integers is 32768 to 32767 Therefore when a calibrated pixel was greater than 32767 it was given a value of 32768 In the L1_A2 version the Cal max value was changed to 32767 which correctly handles pixel values greater than 32767 i e they are truncated at 32767 Saturated lac area A modified version of cubesat repairs errors in the translation of saturated pixel locations to band sample and frame numbers leila modified version accurately reports band and frame number but the sample number is off by plus 1
51. ew location FOV SWIR 200 8 201 8 169 23 99 92 94 93 190 113 203 115 116 138 165 148 119 240 120 240 168 256 VNIR 1 35 1 Spatial co registration of the VNIR and SWIR data is the last step in Level 1_B data processing SWIR data are shifted into alignment with the VNIR pixels see Figure 11 In the cross track field of view FOV dimension X the SWIR pixels are shifted by 1 FOV pixel 1 is removed 1 and pixel 256 is padded with zeros 2 In the along track dimension Y FOV pixels 129 256 are shifted by 1 pixel For frame 1 FOV pixels 129 256 are padded with zeros 3 and the last frame is removed 4 A metadata file EO1YYYYDDD VVVVSSSS rl XXX aln log is generated to show that co alignment has been performed on the level 1 data product The Level 1_B saturated pixel report EO1YYYYDDD_VVVVSSSS_rl_XXX sat contains the location of both LO and L1_B saturated pixels the co alignment process changes the locations of SWIR pixels X 1 256 co aligned image Y
52. filenames have an extension attached that indicates the intermediate Level 1 process e g after smear correction echo correction dark subtraction from which the file was generated although the intermediate files are not included on the tape The metadata file ID number is provided for historical reference and is not required for the user However it is the number referenced in the flow diagrams presented in figures 12 and 13 The L1 appendage is reserved for the final fully processed radiometric calibrated output of the Level 1 code The L1_A appendage is reserved for the final fully processed radiometric calibrated output of the Level 1_A code The L1_B appendage indicates the final fully processed radiometric calibrated and spatially co registered output of the Level 1_B code The _rX in the file name is a revision number that allows differentiating files that have been processed more than once for some reason e g revised calibration file or subsetting for a specific application The file with the L1 hdr Ll A hdr or L1_B hdr appendage is an ENVI ready header for linking the band number with the band center wavelength The center wavelength for field of view 128 is supplied for this header file This header file should be used with caution It is a useful tool for quick looks but the full spectral calibration file SpectralLO_revA should be used for any detailed analysis Log files generated during Level 1 processing provide a
53. gt Repair known bad pixel gt Shift SWIR pixels into Echo removal Provide averaged pre gt Log file generated MD10 alignment with VNIR gt Remove echo from smear image MD3A and post xOutput level 1_A file in signed gt Output level 1_B file in signed corrected files image MD8B dark files integer format L1_A integer format L1_B gt Both dark and image files gt Compute average value gt Generate alignment log file gt Log file generated MD8 of corrected pre andiposti J saeericectent ope cet eee i MD17 image dark files Log file rae 1A EE Hohe I generated reporting f i i average values MD5A for Step7 andrroves rigt to Step8 pre image and 5B for post y image A Step 6 QA L1 B image Step 9 Tar files to DLT gt index of tape contents MD16 gt ancillary data Step 10 HYP_PREFLIGHT_RX doc MD1 gt Display image in ENVI i Hypgain bin MD7 gt Complete QA form during Ship to GSFC for SpectralLO_revA 1 dat MD13 evaluation MD11 distribution to BandWidthL0_revA 1 dat MD14 gt Level 1 data in subdirectories EO1XXX hyp downlink sceneid L1_B MD2 MD3A MD3B MD5A MD5B gt Add center wavelengths to ENVI header hdr file MD12 Science Team MD8 MD9 MD10 MD11 MD12 MD15 MD1 35 EO 1I Hyperion Science Data User s Guide HYP TO 01 077Rev Public Release L1_B Table 8 Hyperion Level 1 Data Set Ll processed data using Level 1 code Ll A processed d
54. he process takes about 15 minutes with the longest time being the 10 minute wait after the instrument is commanded from Idle to Standby to allow the ASPs to achieve thermal stability before the pre image dark and the 3 minute wait after the internal calibration lamp is turned on to allow the lamp output to stabilize The instrument is commanded into Imaging from Standby mode and returns to Idle mode after the post lamp dark collect The instrument is in Standby mode only during DCEs The remainder of the time the instrument is in Idle mode The instrument returns to Idle mode after the post lamp dark collect Further details of the DCE timeline are discussed in chapter 3 The Hyperion science data and ancillary data obtained during a DCE is stored on the Wideband Advanced Recorder Processor WARP The Hyperion science data includes the 5 data files described above and the ancillary data refers to the instrument telemetry obtained during the DCE The WARP also stores science data from the other instruments as well as ancillary data from the other instruments and select subsystems The science data and ancillary data are downlinked to one of several ground stations using an X band downlink All of this data is recorded on Ampex tape and sent to GSFC by the ground stations The data flow at GSFC is shown in Figure 5 GSFC performs Level 0 processing on the data Description of this processing is provided in chapter 3 GSFC sends the Level 0 data and ancillary
55. ing codes use a dark value that is interpolated in time between the pre image dark and post image dark although the Level 1_B code includes the full wait time between the pre image dark collect and the start of the image when performing dark interpolation calculations A calibration file HypGain is applied to radiometrically correct the images An improved calibration file was released at the same time as the release of the Level 1_A code HypGain was used by the Level 1 code and HypGain_revA was used by the Level 1_A code For the current revision status of the HypGain file and as brief summary of changes please refer to the 31 EO 1I Hyperion Science Data User s Guide HYP TO 01 077Rev Public Release L1_B Lireadme_r3 file found in Appendix 1 The HypGain file is used for all Hyperion scenes This file is described in more detail in chapter 4 Pixels that are known to be dead are replaced with an average of neighboring field of view FOV pixels or with 100 of the neighboring pixel if on the edge of the FOV prior to generating the Level 1 data file The bad pixel list was defined pre flight A few bad pixels have been added since the pre flight list was generated These pixels are listed in Appendix 1 Some of the pixels on the bad pixel list are in the portion of the focal plane which is not calibrated The remaining pixels that are corrected for the user are listed below and presented in terms of Spectral Channel Band field of vi
56. ity check is a qualitative assessment performed using ENVI as a visualization tool The band animation feature in ENVI is used to quickly examine all of the bands in the segment of the image being assessed Any unusual bands are then examined in more detail Several specific qualities are checked Streaking indicating residual unevenness across the FOV pixels Banding indicating unevenness as the image is scrolled along the time dimension Shading indicating responsivity roll off at the edges of the image Saturation indicating existence of saturated pixels Focus indicating any out of focus areas on either focal plane Linedrops indicating a line or set of lines in the vertical FOV or horizontal time directional where pixels were temporarily dead or hot known bad pixels are repaired In general streaking and linedrops are the only characteristics that are noticed Streaking can occur in the VNIR or the SWIR It is a result of slight variations in the pixel to pixel calibration in the cross track direction This is discussed further in chapter 4 The linedrops is generally limited to cases in which it is believed that the South Atlantic Anomaly influences Hyperion There are some cases in which saturation may occur For the SWIR this may occur for very hot targets such as gas flares or active volcanoes The VNIR has been seen to saturate on rare occasion in scenes with very bright clouds The saturation limit for the VNIR and SWIR h
57. ive levelO0 ground EO012000147 O1CDO1CC r1 LO Type Hyperion Dimensions 256 pixels x 242 bands x 660 frames Data type 16 bit unsigned integer Byte order big Pixel order BIL Output file File name drb1l2 archive levell ground E012000147_01CD01CC_r1 smear Type Hyperion Dimensions 256 pixels x 242 bands x 660 frames Data type 16 bit unsigned integer Byte order big Pixel order BIL Source drb12 archive level0 ground E012000147_01CD01CC_r1 L0 Start band 71 KKKKKKKKKKKKKKKKKKKK Figure 17 Example MD 9 Smear Removal Log The Pixel Repair Log MD10 has the same initial format as the preceding metadata files but it reports on the repair of known bad pixels The specific file used to identify the known bad pixels is identified badpix txt The results provided in this metadata file are in two parts The first part Bad Pixels identifies the known bad pixels by band or spectral pixel number and sample or FOV pixel number and is essentially a printout of badpix txt This file will be updated over the course of the mission The second Results section indicates how the values of these known bad pixels are replaced For FOV locations other than 1 or 256 the value is replaced by an average of the value in the neighboring FOV pixels e 8 50 of 61 92 and 50 of 61 94 for bad pixel 61 93 FOV pixels 1 and 256 are replaced by the value of the neighbor pixel 2 or 255 42 EO 1 Hyperion Scienc
58. m long along track region Each pixel covers an area of 30 m x 30 m on the ground and a complete spectrum covering 400 2500 nm is collected for each pixel Since Hyperion is a pushbroom system the entire 7 65 km wide swath is obtained in a single frame The 30 m size in the along track direction was obtained by basing the frame rate on the velocity of the spacecraft for a 705 km orbit Hyperion has a single telescope and two spectrometers one visible near infrared VNIR spectrometer and one short wave infrared SWIR spectrometer The Hyperion instrument Figure 2 consists of 3 physical units 1 the Hyperion Sensor Assembly HSA 2 the Hyperion Electronics Assembly HEA and 3 the Cryocooler Electronics Assembly CEA The HSA Figure 3 includes subsystems for the telescope internal calibration source the two grating spectrometers and the supporting focal plane electronics and cooling system The telescope images the Earth onto a slit that defines the instantaneous field of view which is 0 624 deg wide i e 7 65 Km swath width from a 705 Km altitude by 2 44 x10 deg 30 meters in the satellite velocity direction This slit image of the Earth is relayed to two focal planes in the two grating spectrometers A dichroic filter in the system reflects the spectral region from 400 to 1 000 nm to a VNIR spectrometer and transmits the region from 900 to 2500 nm to a SWIR spectrometer The HEA contains the interface and control electronics for
59. nd 24 113 50 of 112 and 114 8 50 of 7 and 9 8 50 of 7 and 9 115 50 of 114 2 2 2 2 2 2 2 2 2 2 3 3 3 3 3 3 6 7 9 9 o o N 30360 pixels fixed out of 40888320 0 074251 KKKKKKKKKKKKKKKKKKK Figure 18 Example MD 10 Pixel Repair Log 44 EO 1 Hyperion Science Data User s Guide HYP TO 01 077Rev Public Release Ll B KKKKK hyp_co_align KKKKK HIP 1 1 Thu Nov 15 17 31 21 2001 Arguments source file EO1E51b hyp 20013071759_SGS E0O10100602001307111PP E012001307_6A8D6A8C_r1_SGS_01 fix S QUEput file EO1E51b hyp 20013071759_SGS E010100602001307111PP E012001307_6A8D6A8C_r1_SGS_01 L1_B output cubename Hyperion L1 Source file File name EO1E51b hyp 20013071759_SGS E010100602001307111PP E012001307_6A8D6A8C_r1_SGS_01 fix Type Hyperion Dimensions 256 pixels x 242 bands x 6702 frames Data type 16 bit integer Byte order big Pixel order BIL Output file File name EOLE51b hyp 20013071759_SGS EO10100602001307111PP E012001307_6A8D6A8C_r1_SGS_01 L1_B Type Hyperion Dimensions 256 pixels x 242 bands x 6702 frames Data type 16 bit integer Byte order big Pixel order BIL Source EOLE51b hyp 20013071759_SGS EO10100602001307111PP E012001307_6A8D6A8C_r1_SGS_01 fix Figure 19 Example MD 17 VNIR SWIR Co registration Log The VNIR SWIR Co registration Log MD 7 has the same format as the Smear Removal Log MD9 4 HYPERION DATA CUBE DETAILS This chapter reviews in
60. nm e SVVIR Band 150 1648 96 nm SVVIR Band 225 2405 63 nm 0 5 5 0 0 7 Wavelength Relative to FOV 128 nm o 0 50 100 150 200 250 Pixel Field of Vievv Figure 23 Variation of the SWIR center wavelength across the field of view 50 EO 1 Hyperion Science Data User s Guide HYP TO 01 077Rev Public Release Ll B 4 3 2 Verification of Spectral Calibration Files A data collection and analysis process to validate the spectral calibration of Hyperion from space was developed The process was based on a solar data collect and an atmospheric limb data collect in which the rays of the sun passing through the atmosphere and reflecting off the Hyperion cover were used The results for the SWIR and VNIR wavelengths confirm that the Hyperion pre flight spectral calibration is valid for on orbit operations As a result the pre flight spectral calibration defined by SpectralLO_revA and BandwidthLO were approved for on orbit operation The revision for the center wavelength was based on a revised analysis of the pre flight data 4 3 3 Spectral Wavelengths Selected for Absolute Radiometric Calibration As mentioned earlier the size of the Hyperion Level 0 and Level 1 data set has the dimension 256x6925x242 However of the 242 spectral channels 204 channels are selected for calibration The reduction is partly due to insufficient signal at the extremes of the spectral range and pa
61. ocessing was performed on June 5 2000 The next set of information contains the command line arguments as well as file parameters associated with the input source and output files The file parameters include the file name file type dimensions data type byte order big endian and pixel order BIL or BSQ are given for each input and output file The file name includes the pathname in the level processing system When appropriate the metadata file will include command line options The end of a log file contains processing Results or Summary information In this case the file is reporting the average dark pixel value for the dark file used in the dark subtraction process The subtracted dark file is 256 FOV pixels x 242 spectral pixels in size and averaged over 220 frames The average dark file is supplied as MD3 The average dark pixel value reported in MDS represents the average value over all FOV and spectral pixels and is used on the Quality Assessment form MD11 to track instrument performance MDS5 EO12000147_01D101D0_r1 avg log KKKKK cubeavg KKKKK HIP 1 1 Mon Jun 5 13 06 10 2000 Arguments source file drb12 archive levell ground E012000147_01D101D0_r1l echo source type hyperion output file drb12 archive levell ground E012000147_01D101D0_rl avg output type generic verbose Source file File name drbl2 archive levell ground EO012000147 01D101D0 rl echo Type Hyperion Dimensions 25
62. od for on orbit radiometric verification 1 4 Summary of Hyperion Performance Characteristics The Hyperion instrument team verified instrument performance during the first three months of the EO 1 mission The assessment focused on determining whether the pre flight Hyperion characterization was still applicable to on orbit operations In addition science data was reviewed in detail to quantify the impact of the instrument characteristics on user applications The instrument performance was compared with requirements and pre flight measurements Tables 1 3 present summaries of the instrument performance and include Radiometric Performance Spectral Performance and Image Quality Performance respectively Each table contains the pre flight value and the on orbit value The EO 1 Hyperion Early Orbit Checkout Report Part H On Orbit Performance Verification and Calibration documents the analysis 10 EO 1 Hyperion Science Data User s Guide HYP TO 01 077Rev Public Release Ll B The on orbit assessment concluded that the pre flight characterization was valid for on orbit operation The pre flight absolute calibration file was adjusted for pixel to pixel variations based on an updated analysis of pre flight results and accepted for on orbit operation The spectral calibration was verified for on orbit operation Table 1 presents the results from the radiometric performance evaluation The signal to noise ratio SNR exceeds the requirement for
63. ode was updated to Level 1_B in response to a request from NASA GSFC to incorporate co registration of the VNIR and SWIR data Processing with the Level 1 B code began in November 2001 1 1 Document Scope The goal of this document is to assist the user in most effectively exploring the Hyperion data set The document is organized into four chapters This chapter Chapter 1 Introduction provides a general overview of the EO 1 Mission an overview of the Hyperion instrument and a review of the Hyperion instrument requirements This chapter also introduces the users to the Hyperion data collection event DCE sequence and highlights the flow of the data from the spacecraft to the users A contact list is provided further support Chapter 2 Hyperion Data Cube Quick Start provides the user with some quick steps to allow the user to become familiar with the data The discussion assumes the user is using ENVI and that the 6 EO 1 Hyperion Science Data User s Guide HYP TO 01 077Rev Public Release Ll B user is reviewing the Hyperion Level 1 processed data having a L1_B extension also applicable to Ll and Ll A extensions The chapter recommends VNIR and SWIR bands for grayscale and RBG quick look images The chapter introduces the user to the Hyperion spectra A brief discussion is provided in the last section for those who may use other software for processing and data analysis The full discussion of how to use the Hyperion data set is deferred
64. places a bad pixel with the average of the two adjacent FOV pixels If a dead pixel is on the edge of the FOV sample 1 or 256 then 100 of the adjacent pixel sample 2 or 255 value replaces the dead pixel Saturated pixels Not reported eee Offset removal dark The dark subtraction is performed using an average of all frames of a dark file that was acquired closest in time to the image cal VNIR SWIR co Not included alignment L1 A Ratio file The ratio revA txt is an improved estimate of the correction factor in regions where signal levels in the data were very low Hypgain_RevA dat adds bands 5 8 400 430 nm to the L1_A data for a total of 204 calibrated bands The agreement between bands of the overlap region 890 925 nm has been improved The dark subtraction process used in the derivation of the pre flight calibration file was revised Radiance values in HypGain_revA dat are generally 1 lower than hypgain txt except below 550 nm and between 900 1100 nm where radiances are more than 1 lower than hypgain bin Calibration multiplier The calibration multiplier was changed to retain precision of the maximum radiance 750 W m2 sr um VNIR and 350 VNIR SWIR W m2 sr um SWIR within the range of a 16 bit signed integer 32768 32767 The calibrated radiance values are multiplied by 40 for VNIR bands 1 70 and by 80 for SWIR bands 71 242 Minimum and maximum radiance values were changed to correspond with the range of a 16 bit signed in
65. rack direction is a little more complicated The difference is dependent on the field of view location The in track difference is zero for field of view pixel 1 The in track difference decreases linearly to 1 for field of view pixel 256 5 END TO END MEASUREMENT ACCURACY Table 9 shows the absolute and precision errors for a single measurement of a scene element by Hyperion The top of the atmosphere radiance measurement error of a scene by a given single pixel at a given spatial location and spectral wavelength is the result of the combination of an absolute bias systematic error and a precision error Table 9 Hyperion End to End Measurement Error Total Measurement Error VNIR SWIR 2 95 3 39 VNIR SWIR Precision Error from Precision table 1 60 2 30 Absolute Systematic Bias 2 49 2 49 52 EO 1I Hyperion Science Data User s Guide HYP TO 01 077Rev Public Release L1_B Appendix 1 Llreadme r3 Effective L1 file Ratio file VNIR SWIR Cal Cal Bad Description of modification Date Cal Cal Max pixel ul Mult file 1 1 Apr 01 01 L1 fratio txt E bin 00 100 0 65536 ae See ae pages 01 HypGain revA dat See 32768 E 12 Oct 01 L Al Joo revA dat ii 32768 32768 ae following pages 16 ES 01 J L1 A2 RE nn revB dat mi ee 32767 badpix3 See following pages 23 ial 01 L1_A3 ratio revB txtiHypGain revB dat rr 32768 132767 badpix3 See following pages 15 Nov 01 EES HypGain revB da
66. rtly due to existence of an overlap region between the VNIR and SWIR The bands that are not calibrated are set to zero The user will notice a signal in the Level 0 data and will see a value of zero for those bands that are not calibrated in the level 1 processed data The range of calibrated spectral channels is different between HypGain and HypGain_revA and is listed in the table below HypGain HypGain_revA VNIR bands nm 9 57 437 925 5 57 400 925 SWIR bands nm 75 225 890 2400 75 225 890 2400 Calibrated channels 200 4 overlap 204 4 overlap Unique Channels 196 200 This results in 204 calibrated channels with 200 unique wavelengths provided in the L1_A data file Four bands in the VNIR 54 57 overlap with four bands in the SWIR 75 78 which is 892 to 926 nm The selection of bands for science applications is left to the discretion of the user 4 3 4 Quick Look Spectral Wavelength in the Header File The level 1 data product includes in the header file a center wavelength and a full width half maximum value for a Gaussian shape The values are applicable for field of view pixel 128 It can be used for quick look plots but should not be used for science applications 4 4 VNIR SVVIR Spatial Alignment Advanced Topics Level 1 and Level 1_A Data Products Only 4 4 1 Ground Sample Distance The ground sample distance for the VNIR and SWIR focal planes was determine
67. s suggested for quick viewing of the Hyperion data Example images are provided after the discussion It should be noted that unless you are using level 1_B data there is a spatial offset between the VNIR and the SWIR So unless the appropriate shifts are made RGB images should be limited to only VNIR bands or only SWIR bands 2 2 1 VNIR Band 40 or SWIR Band 93 A simple and reliable way to get a quick feel for the contents of the image is to display a gray scale image of Band 40 This band in the VNIR corresponds to 753 nm SWIR band 93 at 1074 nm can be used for the SWIR Figure 6 is an example of Band 40 and Band 93 for the San Francisco Salt Pond data collection event obtained on January 17 2001 The entire swath width is displayed but only a subset of the swath length is presented in these images When reviewing single bands streaks in the vertical direction may be noticed The phenomena are caused by a variation in the calibration of the pixels in the cross track direction Small pixel to pixel variations are more noticeable in uniform scenes This topic is discussed further in chapter 4 15 EO 1 Hyperion Science Data User s Guide HYP TO 01 077Rev Public Release Ll B 753 nm VNIR Band 40 1074 nm SWIR Band 93 Figure 6 Example Gray Scale display of Hyperion data San Francisco Salt Ponds 1 17 01 2 2 2 VNIR visible RGB To obtain a color image of the scene that represents true RGB the bands 29 23 16 for R G B are typically u
68. s verify the ground measurement to within the measurement accuracy for both VNIR and SWIR Table 2 Spectral Performance Instrument Pre Flight On Orbit Parameter Number of VNIR amp 220 220 Spectral Channels SWIR 200 selected for Level 1 processing Spectral Range 357 2576 nm 357 2576 nm center wavelengths 400 2400 nm selected for determined to 1 nm Level 1 processing 11 EO 1 Hyperion Science Data User s Guide HYP TO 01 077Rev Public Release Ll B Instrument Pre Flight On Orbit Parameter Spectral VNIR 10 08 10 09 nm Not measured Bandwidth di dl SVVIR 10 11 10 13 nm Not measured Cross Track VNIR 2 57 3 59 nm 1 71 2 55 nm Spectral Error lt SWIR 17 08 nm 40 97 nm Table 3 presents results for the Image Quality parameters The measured Ground Sample Distance GSD at nadir was 30 meters as predicted from pre flight measurements There are 256 field of view locations that comprise the swath width This corresponds to a 7 75 km swath width per focal plane However there is a 1 pixel cross track pixel difference between the VNIR and SWIR which reduces the swath width by one GSD 255 pixels used The VNIR SWIR Modulation Transfer Function MTF was measured on orbit and the results were similar to the pre flight measurements Spatial co registration is a measure of an object s position as a function of spectral band within the focal plane The characteriza
69. sed This band combination corresponds to approximate wavelengths of 641 nm 580 nm and 509 nm Slight variations in the bands selected will not noticeably affect the RGB image Figure 7 displays an example of a visible RGB for the San Francisco Salt Pond data collection event obtained on January 17 2001 16 EO 1 Hyperion Science Data User s Guide HYP TO 01 077Rev Public Release Ll B 641 380 509 nm 855 580 509 nm 2194 1649 1074 nm 53163 10 23 16 0413133 VNIR Visible RGB VNIR Vegetation RGB SVVIR RGB Figure 7 Example of RGB vievving of Hyperion Data San Francisco Salt Ponds 1 17 01 2 2 3 VNIR vegetation RGB To obtain a color image of the scene in which vegetation appears red bands 50 23 16 can be selected This band combination corresponds to approximately 855 nm 580 nm and 509 nm Slight variations in the bands selected will not noticeably affect the RGB image See Figure 7 for an example 2 2 4 SWIR RGB To obtain a color image of the SWIR bands 204 150 93 for R G B is a usable combination The corresponding wavelengths are 2194 nm 1649 nm and 1074 nm These bands are outside of the region of the spectrum that is most significantly affected by atmospheric absorption See Figure 7 for an example When using the SWIR data it is important to know if the SWIR was at the proper operational temperature The proper operational temperature is when the SWIR FPE temperature is 153 5 1C An HDF file delivered with the Hyperion
70. so Hyperion views the ground directly For a solar calibration or atmospheric limb collect the spacecraft pitch yaw is modified so that Hyperion views the reflection off the diffuse surface on the inside of the cover In both cases the pointing direction is commanded prior to the scene such that the spacecraft motion has settled prior to the start of the collect For a lunar stellar or planetary collect the spacecraft motion is commanded throughout the collect which results in a scan Table 5 Types of Data Collection Events scene Atmosphere Collect Spacecraft View effects sec Motion Ground Yes 30 Point Direct Lunar No varies Scan Direct Stellar Planetary Solar No 16 Point Diffuse Reflection Atmospheric Yes 16 Point Diffuse Limb Reflection 3 3 Data Flow from the Spacecraft to the User There are a few steps that have to occur in order for the user to obtain Hyperion data for a specific scene For example the user must submit a request to GSFC that of a desired site should be collected The user would supply the latitude and longitude of the site as well as the date of the collect if it were critical The user should also provide any other additional information that may be important to the planners For example day night collect take no matter what take only if no clouds ground truth coincident is some factors that would be worthwhile to relay to the planners The scene would then be sch
71. t Es 32767 badpix3 See following pages 53 EO 1I Hyperion Science Data User s Guide HYP TO 01 077Rev Public Release L1_B Level 1 Processing Change Summary Level 1 version cal L1_A1 L1_A2 L1 A3 LI B Dc DN DL SL teat DL NE DS J DT i o Gantt ran an fosa pS ReGen rence larcaniarar Meca regar ncan eet w e p o A _ o e pe v po po marae Td LE Saturated LISIC I po ves res upsets es sata Yvos pate pos peasy rani coatgmenpo fe eff change from previous version yes included in level 1 version no not included in level 1 version 54 EO 1I Hyperion Science Data User s Guide HYP TO 01 077Rev Public Release L1_B Level 1 Processing Change Record Level 1 Parameter or Description of Change version Algorithm Ratio file file ratio txt was derived from ratio txt was derived from pre flight instrument characterization 0 E EE a P A instrument characterization Sage a aaa Data Type The data type is a 16 bit unsigned integer ranging from 0 65536 CE maximize precision and reduce file size the VNIR and SWIR calibrated radiance values are multiplied by 100 and stored as VNIR SWIR integers Cal Min Max Minimum and maximum calibrated radiance values are based on the range of 16 bit unsigned integers 0 65536 The cal min max values cut off radiance values lt zero at 0 and values gt 65536 are truncated at 65536 Bad pixel repair Cubefix re
72. teger Calibrated radiance values lt 32768 are cut off at 32768 and values gt 32768 are truncated at 32768 Saturated pixels The cubesat program identifies raw LO image pixels with values equal to 4095 and reports their band sample and frame numbers to a log file sat file extension The log file also identifies smear and echo affected pixels because these artifacts can not be removed correctly when the source pixel is saturated The following errors were discovered in the sat file after the release of L1 A data 1 the location of saturated pixels in the image file is incorrectly translated into band sample and frame number 2 smear and echo affected pixels are incorrectly reported for VNIR bands 1 70 since these corrections are not applied to the VNIR data These errors occur in versions L1_A L1 A1 and L1_A2 Offset removal dark Dark signal removal for L1 A data is performed using a dark file value that is interpolated in time between the pre image dark Cal and post image dark files The interpolated values are subtracted from the image on a frame by frame basis EO 1I Hyperion Science Data User s Guide HYP TO 01 077Rev Public Release L1_B Level 1 Parameter or Description of Change version Algorithm L1 Al oe file ratio_revB txt contains zeros at dead pixel locations to prevent the smear correction to be propagated through the echo ae JET at these locations Bad pixel file Bad pixel file file Added three non r
73. the instrument and the CEA controls 8 EO 1 Hyperion Science Data User s Guide HYP TO 01 077Rev Public Release Ll B cryocooler operation These units are all placed on the nadir facing deck of the spacecraft with the viewing direction as shown in Figure 4 HSA Figure 2 Drawing of Hyperion Instrument and Electronics Figure 3 Photo of the Hyperion Sensor Assembly HSA EO 1I Hyperion Science Data User s Guide HYP TO 01 077Rev Public Release L1_B Hyperion Advanced Land Imager Anas phere Corrector Figure 4 Photo of Hyperion as mounted on the EO 1 spacecraft The HSA consists of an enclosure providing thermal control for the Opto Mechanical Subsystem OMS on which are mounted the VNIR and the SWIR Focal Plane Arrays FPAs The OMS consists of the telescope the VNIR spectrometer and the SWIR spectrometer The HSA enclosure is the mounting interface between the HSA and the spacecraft and has a motorized aperture cover The HSA enclosure also provides support for the pulse tube cryocooler the VNIR and the SWIR Analog Signal Processors ASP and the in flight calibration source IFCS The SWIR FPA is cooled by acryocooler The VNIR FPA is cooled by a radiator The IFCS consists of a lamp to illuminate the backside of the aperture cover which is a diffuse reflector white paint In addition with the aperture cover partially open solar illumination of the diffuse reflector provides a second meth
74. tion of the VNIR and SWIR spatial co registration of spectral channels was within the on orbit measurement accuracy In this case the on orbit status is considered to support the pre flight characterization The VNIR SWIR spatial co registration is discussed further in chapter 4 Table 3 Image Quality Performance Instrument Pre Flight On Orbit Parameter GSD nadir Entire Range 29 88 m 30 38 m Swath Entire Range 7 75 km 7 75 km Width per focal plane per focal plane MTF 450 nm 22 29 meas 500nm 23 27 meas 500nm In Track 630 nm 22 27 23521 900 nm 22 24 24 28 1250 nm 27 30 20 25 1650 nm 25 21 28 2200 nm 23 28 Not measured VNIR All 10 25 of pixels Consistent with pre flight spatial Co 10 30 of pixel Registration SWIR All 18 28 of pixels Consistent with pre flight spatial Co 25 15 of pixel Registration 1 5 Overview of Hyperion Data Collection and Data Flow A Hyperion science data collection event DCE consists of 5 files In order of collection this includes pre image dark collect image collect post image dark collect internal calibration lamp collect post lamp dark collect The dark collects are 1 second each the lamp collect is 3 seconds 12 EO 1 Hyperion Science Data User s Guide HYP TO 01 077Rev Public Release Ll B and a typical image collect is 30 seconds with the desired scene being contained in the centered 24 seconds T
75. tions assume different byte orders For example PCs SGI Windows SGI LINUX and Dec Alpha UNIX use little endian byte order while SGI IRIX and Sun UNIX use big endian ENVI is compatible with either and reads this attribute 28 EO 1 Hyperion Science Data User s Guide HYP TO 01 077Rev Public Release Ll B Level 0 File Generated At Frame Numbers Dimensions Number of Cross Track Pixels Number of Bands Number of Along Track Pixels Pixel Order Number of Missing Pixels VNIR Sync Time beginFrame endFra me syncTime SWIR Sync Time beginFrame endFra me syncTime VNIR Time Code beginFrame endFra me TC3 TC4 TC5 TC provides the time at which the level 0 processing was performed in year yyyy Julian day ddd and GMT hhmmss provides the number of the start and stop frame for the file The start frame number is 0 For a nominal DCE image collect the end frame number would be 6925 For a dark file the end frame number would be 223 are given as the number of cross track pixels x number of spectral bands x number of frames along track pixels The number of cross track or field of view FOV pixels is fixed at 256 This number will not vary The number of bands is the same as the number of spectral channels which is 242 and does not vary Some images may not have valid SWIR data if the SWIR focal plane was not held at its operational temperature during the imaging event but there will
76. to chapter 4 Chapter 3 Hyperion Data Collection Flow and Processing describes the data collection event sequence and the types of data collection events that are available The data flow from user request through shipment of data to the user is highlighted Further details of the level 0 and level 1 processing are discussed For the level 1 L1 L1_A and L1_B processing discussion the steps as well as the data files that are created at each step of the processing sequence are presented Chapter 4 Hyperion Data Cube Details describes the derivation and verification of the absolute radiometric calibration and spectral calibration The section on advanced topics presents additional details The details are required to properly use the absolute and spectral calibration Chapter 5 End to End Measurement Accuracy describes the Hyperion end to end error budget 1 2 EO 1 Mission Overview The responsibility of NASA s Mission to Planet Earth is to ensure the continuity of future Landsat data and the improvement of Earth Science information for better understanding of our planet The New Millennium Program s NMP Earth Orbiter 1 EO 1 serves both of these responsibilities EO 1 Advanced Land Imager ALI validates technologies which could provide cost reductions for future Landsat missions and the Hyperion instrument provides a new class of earth observation data for improved Earth surface characterization For the latter the primary go
77. to identify saturated pixels as well as those pixels whose value is suspect because of invalid echo and smear corrections Timed with the release of the Level 1_A code was a release of an improved calibration file and improved echo correction file There was one significant revision associated with the Level 1_B code Level 1_B follows the same processing steps as Level 1 A but in addition Level 1_B spatially co registers the VNIR and SWIR data In addition the Level 1_A log file sat that reports on saturated pixels in the level 0 data was modified to indicate the location of the pixels in the Level 1_B data because the co registration process shifts the SWIR pixel locations Finally the Level 1_A dark interpolation process was modified slightly to account for the full duration of time between the pre image dark file and the start of the image 30 seconds GSFC sends the Level 0 processing data to TRW along with a request list of scenes to be processed to Level 1 TRW performs the Level processing A flow chart for the Hyperion Level processing using the original Level 1 code is shown in Figure 11 A flow chart for the updated Hyperion Level 1_ A and Level 1_B processing is shown in Figure 12 The charts are broken into data processing steps and indicate the function that is performed and the metadata file that is created to track the processing steps Ground image solar calibration and lunar calibration data are processed identically
78. tric calibrated Hyperion science data Hyperion level 1 processing applies to processing performed using the Level 1 code Level 1_A and Level 1_B code Processing with the Level 1_A code began July 1 2001 Processing with the Level 1_B code began November 15 2001 The four revisions in the Level 1_A code include 1 Instead of subtracting the dark file nearest to the image an interpolated dark file is calculated using the dark files collected pre and post image and subtracted from the image 2 Output of the level 1_A code is a signed integer rather than an unsigned integer as in level 1 3 Level 1 output was in units of watts sr micron m7 x 100 Because of the change listed in 2 above the multiplication factor was revised so that VNIR is in units of watts sr micron m x40 and SWIR is in units of watts sr micron m x80 This change allows retention of maximum precision without overflowing the size limit for 16 bit signed integer output 4 Inthe unusual event that pixels are saturated in the SWIR the smear and echo corrections become invalid This has been observed in only a few scenes out of the 1400 collected to date 30 EO 1 Hyperion Science Data User s Guide HYP TO 01 077Rev Public Release Ll B and are typically gas plumes or volcanic hot spots Because saturation affects the validity of the data in subtle ways the Level 1_A code first scans the level 0 data for saturated pixels and produces a log file sat

EO-1/ Hyperion Science Data User's Guide, Level 1_B

Contents

Download Pdf Manuals

Related Search

Related Contents