CLoud Archive User Service User Guide - BADC
Contents
1. The Quicktime Player software that runs on Macintosh computers enables the animations to be viewed in a flexible manner for example between selected date time limits or in continual loops The animations can also be paused and stepped backwards and forwards using the left and right arrow keys To view the animations under Unix you may need to download and install suitable software for example XAnim http smurfland cit buffalo edu xanim home html The animations run best if the greyscale option Cg is used The animations may also be viewed on Windows PCs but you may need to download and install the PC version of Quicktime from the Apple Web site at http www apple com quicktime go to the download section A3 Bibliography Hodges K I 1996 Spherical non parametric estimators and their application to the UGAMP AMIP integration A ten year cyclonic climatology for the northern and southern hemisphere winters Mon Weather Rev 124 pp2914 2932 Hodges K D W Chappell G J Robinson and G Yang 2000 An improved algorithm for generating global window brightness temperatures from multiple satellite infra red imagery J Atmos And Ocean Tech 17 1296 1312 Rossow W B A Walker and M Roiter 1997 International Satellite Cloud Climatology Project ISCCP Description of Reduced Resolution Radiance Data WMO TD No 58 World Meteorological Organization Geneva Salby M L H H Hendon K Woodberry and K2. CLoud Archive User Service User Guide Version 1 4 Authors Gary J Robinson and Kevin I Hodges NERC Environmental Systems Science Centre The University of Reading Whiteknights Reading Berkshire RG 6 6AH Copyright 2005 NERC Environmental Systems Science Centre 1 OVERVIEW 2 THE DATA 2 1 2 2 2 3 2 4 2 5 2 6 3 DATA PROCESSING 3 1 3 2 3 3 3 4 3 5 4 ACKNOWLEDGEMENTS 5 ACCESS TO THE ARCHIVE AND CONDITIONS OF USE APPENDICES A1 Contents Introduction File Naming Convention Image File Organisation PGM Image File Format Data Scaling and Geographical Arrangement Quality Data Files 2 6 1 Contributing Satellites 2 6 2 Interpolation Quality The ISCCP B3 Data Processing Methodology Limb Darkening and Cloud Geometry Correction Quality Control of the ISCCP B3 Data Accuracy of the CLAUS archive Portable Grey Map File Format A2 Monthly Animations A3 Bibliography A4 Relevant Links CLAUS Project CLAUS Consortia Members Other Links Ilear Giida 1 A 1 Overview The CLAUS project was supported by the European Union under the IVth Framework Programme Environment and Climate and ran from April 1997 to December 1999 The objectives of the CLAUS project were to produce a long time series of global thermal infra red window 10 5 12 5 um imagery of the Earth and to test the feasibility of using this in validating atmospheric General Circulation Models GCM
3. Tanaka 1991 Analysis of global cloud imagery from multiple satellites Bull Am Met Soc 72 467 480 Yang Gui Ying Julia Slingo 2001 The Diurnal Cycle in the Tropics Monthly Weather Review 129 No 4 784 801 Ilear Giida 1 A 15 J J C Cloud Archive User Service CLAUS Cheruy F R S Kandel and J P Duvel 1991 Outgoing longwave radiation and its diurnal variations from combined Earth Radiation Budget Experiment and Meteosat observations 2 Using Meteosat data to determine the longwave diurnal cycle Journal of Geophysical Research Atmospheres 96 D12 22623 22630 A4 Relevant Links CLAUS Project Web Server http www nerc essc ac uk CLAUS CLAUS Consortia Members ESSC http www nerc essc ac uk UGAMP http ugamp nerc ac uk CNRM http www meteo fr LMD http www lmd ens fr MPI http www mpimet mpg de ECMWF http www ecmwf int Other Links BADC http www badc rl ac uk EOSDIS http spsosun gsfc nasa gov eosinfo EOSDIS _Site index html ISCCP http isccp giss nasa gov Hadley Centre http www meto govt uk research hadleycentre Ilear Giiida 1 A 1A
4. moves preferentially to higher cooler layers The total radiation received is thus less than that received when viewing from nadir and hence appears to emanate from a cooler body i e it has a lower brightness temperature Left uncorrected limb darkening would lead to apparent discontinuities when the data from the satellites are fused together To counteract this a simple empirical correction is applied L z L O a b In cos z 1 where L is radiance z is zenith angle and a and b are empirically determined coefficients This correction is applied to all satellites geostationary and polar orbiters Another zenith angle dependent effect on observed brightness temperature is caused by the differing viewing geometries of the geostationary and polar orbiting satellites At high zenith angles the chance of a geostationary satellite viewing through broken cloud is considerably reduced In contrast a polar orbiting or an adjacent geostationary satellite viewing the same scene but closer to the nadir stands a greater chance of seeing through the cloud to the warmer surface below To reduce this cloud geometry effect the data from both geostationary and polar orbiting satellites are down weighted by a function similar to 1 Clearly in regions where data from only one satellite is available this correction has no effect 3 4 Quality Control of the ISCCP B3 Data The quality control procedures applied to both the ISCCP B3 thermal in
5. or fields Z I and M Table 3 I Z 615 4 3 2 1 0 M 7 Table 3 Structure of Interpolation Quality byte Field Z contains the mean cosine of satellite zenith angle linearly scaled to span the range 1 0 gt 0 1 Field I contains the sampling interpolation level with the values listed in Table 4 Field M is the missing data flag 1 missing 0 present If M is set then the other fields are undefined The three fields are arranged so that the higher the combined byte value the lower the reliability of the corresponding Tg value Strictly speaking M is redundant in the iq file since this information is available in the bt file as value 0 However it is included in the iq file for completeness Ilear Giida 1 A A hive l Value Interpolation level 0 primary spatial secondary spatial tertiary spatial 38 primary temporal secondary temporal 3 4 6 7 reserved 7 reserved Table 4 Values of Interpolation levels The Z values are generally only indicative since strictly speaking they are only defined for brightness temperature values obtained from a single satellite Although the cosine of zenith angles are always positive for valid observations so that mean of more than one value is also always positive no account is taken of the distribution of the observation azimuth angles The iq data are useful when analyses are performed on the brightness temperature data since there are o
6. 12 5 wm denoted bands 4 and 5 In the ISCCP B3 dataset only band 4 is calibrated this band and the nominal 10 um window channel from the geostationary satellites are used in producing the CLAUS dataset The combined imagery from the geostationary satellites usually provides continuous coverage up to 75 latitudes whilst the polar orbiter imagery covers higher latitudes and is also used to fill in where geostationary coverage is poor e g the Indian Ocean region and where data from individual geostationary satellites are absent or incomplete The continual sampling scheme of the polar orbiters requires that these data have to be processed in such a way as to reduce their contribution with increasing time away from the synoptic times 0 3 6 etc hours 3 2 Processing Methodology The approach used in creating the data for the CLAUS archive was necessarily pragmatic due to the demands of the CLAUS project This involved implementing a two step procedure comprising an initial spatial interpolation of the source image data followed by temporal interpolation to fill in gaps arising from missing source images parts of images or regions covered only by polar orbiting satellites i e the poles Spatial interpolation onto the uniform latitude longitude target grid is performed using spherical kernel estimators Hodges 1998 at three levels using different kernel widths 0 5 primary 1 0 secondary and 1 5 tertiary The level at each
7. 92 at 21 00 UT The leftmost column corresponds to the Greenwich Meridian and is repeated at the rightmost column Tg values are stored using an inverted greyscale section 2 5 so that clouds appear bright Llear Gitida 1 A 2 CLAUS Ancillary images provide an indication of the quality of the Tg data on a per pixel basis and are held as two separate PGM image files for each three hourly Tg image One file acts as an index to the satellites contributing to the Tg image whilst the other denotes the type of interpolation used in deriving the Tg values and the representative satellite look angles section 2 6 2 2 File Naming Convention Each image file in the archive is named YYYYMMDDHH 2xx where YYYY is the year MM is the numeric month January 01 DD is the calendar day number and HH 00 midnight 03 06 09 12 15 18 and 21 is the nominal Universal Times UT The xx part of the file extension denotes the data type bt Brightness Temperature es Contributing Satellites and iq Interpolation Quality The 2 in the file extension indicates that the data refer to ISCCP channel 2 i e the thermal infra red window channel Note that the one half and one third degree resolution versions of the data have the same naming convention They are therefore held in separate directory trees the assumption is that users will work with data at only one resolution The Quicktime monthly animation files Appendix A 2 are named YYYYM
8. C web site You will then be given access to the data from an account on one of the BADC computers In all publications relating to work using data from the CLAUS archive you are requested to include an acknowledgment of the form These results were obtained using the CLAUS archive held at the British Atmospheric Data Centre produced using ISCCP source data distributed by the NASA Langley Data Center If the data provided a significant basis of the work co authorship would be appreciated Ilear Giiida 1 A 12 Bu Cloud Archive Usi Appendices A1 Portable Grey Map File Format uo Type BT CLAUS Brightness Temperature Image Data Resolution 0 5 Half degree Synoptic Date 1994061315 Source Channel 2 TIR Satellites 13 63 54 32 43 44 00 00 Creation Date 2001 10 03 13 41 26 Se Sk Se Se Sk Se SR HD HEADER Revision 4 50 ESSC 720 359 255 lt width gt X lt height gt unsigned byte data values in sequence starting at the top left corner of image and proceeding row wise DATA Figure 5 The Portable Grey Map file format used in the CLAUS archive This example is a brightness temperature image for 1994 06 13 at 15 00 hours at one half degree resolution For one third degree resolution data the resolution value is 0 3333 and the width and height are 1080 and 539 respectively Explanation Line 1 is the PGM magic number denoting data are binary bytes Lines 2 8 are PGM comment line
9. M_2 mov and are stored in the movies directory of the CLAUS archive under annual sub directories 2 3 Image File Organisation To facilitate data access the three CLAUS products are arranged within the archive in a four level directory hierarchy with the top level directory denoting the channel number 2 at present The second level contains the lo_res hi_res amd movies directories The third level contains the calendar year For the image data products only the fourth directory contains the year plus calendar month The file naming convention used ensures that each file name within each of the lo_res or hi_res hierarchies is unique channel 2 lo_res hi_res movies 1983 1986 1999 as lo_res directory 1983 1985 1999 198601 198604 19861 198501 2 mov 198502 2 mov 1986040715 2bt 1986040715 2cs 1986040715 2ig Figure 2 Directory structure of CLAUS archive Should further ISCCP channels be processed these data will be placed in separate directory hierarchies Ilear Guida 1 A Q CLAUS 2 4 PGM Image File Format The archived data are stored using the Portable Grey Map PGM image format for two reasons First they may be viewed directly on screens using proprietary graphics packages or widely available public domain packages e g Image Magik Second converting the data to other formats to suit user requirements is fairly straightforward a sample C programme to con
10. ccasions where the spatial sampling of the source data is not optimal and spurious results may arise see section 3 5 In such cases examination of the iq data values will indicate problem regions for example where the Tg values are temporally interpolated or where zenith angles are large i e near limb observations Figure 4 Interpolation Quality information for the Tg image shown in Figure 1 Colours denote the degree of smoothing in the spatial red primary green secondary blue tertiary and temporal magenta primary orange secondary interpolation processes Intensity of the colour indicates the off nadir angle bright nadir Note that the majority of values are interpolated spatially at the primary i e highest resolution Ilear Giida 1 A 7 CLAUS 3 Data Processing This section describes briefly the data processing used by ESSC to create the CLAUS archive For a comprehensive account refer to the publication by Hodges et al 2000 3 1 The ISCCP B3 Data The ISCCP B3 source data for the project consist of two types of imagery whole earth images from operational meteorological geostationary satellites METEOSAT GMS and GOES sampled at synoptic times corresponding to 0 3 6 etc hours UT and continuous time series of imagery from the NOAA polar orbiting meteorological satellites The window channel is split for the AVHRR sensor on NOAA satellites 7 onwards into two channels nominally 10 5 11 5 um and 11 5
11. en the brightness temperature images were viewed as a movie sequence this technique has became one of the main quality control checks for the final product The fact that these errors have been noticed only now is mainly due to the ability of the CLAUS processing system to allow data from multiple satellites to be compared simultaneously To counteract these problems a suite of automated procedures has been developed by ESSC for use during different stages of the data processing The first step is to automatically identify and where possible remove noisy pixels and bad scan lines from each image These are flagged either automatically or with some user intervention in ambiguous cases so that the corresponding data are ignored in subsequent processing Navigation and image distortion errors are then identified automatically by noting images with spurious off planet pixels Initially such images were simply rejected but recently techniques have been developed to perform navigation correction automatically by using the imagery from the other satellites and reference coastline data sets Data generated by National Remote Sensing Centre under contract to ESSC as part of the CLAUS project using the ISCCP source data as is Data generated by ESSC either during the CLAUS project or afterwards with identification and removal of individual images with obvious navigation or calibration errors Generated by ESSC after end of project usin
12. fra red source data and the resulting CLAUS brightness temperature imagery increased in sophistication during the lifetime of the CLAUS project and continue to be improved As a result the data in the archive are of several distinct levels of quality those of a lower quality will be replaced gradually by higher quality data as the better quality control checks are applied to data as they are reprocessed The B3 data are notionally quality controlled by ISCCP 1987 This involves identifying and flagging suspect lines in individual images navigating each image to assign latitude longitude values to each pixel etc However ISCCP and CLAUS have different objectives in terms of products ISCCP aims to generate monthly averaged cloud products on a per satellite basis whilst CLAUS aims to generate three hourly global composite images CLAUS therefore has a far higher reliance on accurate quality control of the B3 data than ISCCP by definition one and half orders of magnitude greater During the course of processing data for the CLAUS archive it became evident that a significant fraction of the B3 data contained residual errors poor image quality noise and mis navigated images mainly from the early GOES series Other errors appear as mis calibrated individual images and apparent differences between sensors from different satellites The effects these Ilear Giiida 1 A a errors had on the original versions of the CLAUS archive were apparent wh
13. g improved limb darkening correction techniques Calibration trend errors identified and corrected Data from geostationary satellites given temporal weighting to improve results in overlap regions Navigation errors in individual images identified and corrected mainly GOES series Identification and correction of individual images containing rectification errors mainly GOES 7 some METEOSAT Identification and correction of individual images containing gross calibration errors Table 5 Version history of the CLAUS archive Note superseded versions of the data are not retained Ilear Giida 1 A 1n 3 5 Accuracy of the CLAUS archive A number of factors determine the accuracy of individual Tg values within a CLAUS image limb darkening and cloud geometry effects time difference between source data and nominal 3 hour time differential absorption by water vapour sampling characteristics of the satellites sensors calibration errors and the interpolation process itself The magnitude of these errors varies and can be difficult to quantify but as a guide they are Limb darkening Atmospheric absorption reduces Tgs by 10 K for clear sky regions near the limb of geostationary satellites and up to 3 K at the edge of polar orbiter swaths After correction the residual errors are estimated to be 1 2 K For cloudy scenes the near limb swath edge error is higher 4 K because of cloud geometry effects Te
14. grid point having the highest data density is chosen to represent the brightness temperature at that point This approach preserves the detail in the source data where present but reduces the need for temporal interpolation in areas of low density Temporal interpolation is then performed on those spatially interpolated three hourly brightness temperature images that have data voids using the two preceding and two succeeding images to estimate Tg values within the data void The level of temporal interpolation as stored in the iq file is determined by which of the images 3 or 6 hours away have valid values in the data void of the image being processed primary both within 3 hours secondary only one within 3 hours otherwise data point is flagged as missing This technique works very well at low to mid latitudes and reasonably well at high latitudes due to the precessing nature of the orbits of the polar orbiting satellites Ilear Guida 1 A R 3 3 Limb Darkening and Cloud Geometry Correction A feature of thermal infrared imagery of the Earth particularly from geostationary satellites is the noticeable darkening of the image towards the Earth s limb so called limb darkening As the zenith viewing angle increases the path length through the atmosphere increases This has two consequences First radiation emitted by or near to the surface is attenuated mainly by water vapour Second the emission by the atmosphere itself
15. mporal mismatch Diurnal variation over cloud free land can be up to 20 K over local Summer scenes up to 10 K over cloudy regions due to cloud motion Differential absorption typically 1 2 K differences in clear sky moist regions between Tps from geostationary and polar orbiting satellites due to the latters narrower pass bands Calibration errors Errors in individual satellites scenes have been found of magnitude 15 K although these are rare and can be easily spotted and corrected Systematic errors of up to 3 4 K are more common these can also be corrected but residual errors in individual satellite images are estimated to be 1 2 K Sampling errors The finite and different footprint sizes of the different sensors means that observations of Tgs over unresolved structures is highly variable estimated to be up to 2 3 K Interpolation errors Same effect as temporal mismatch but can be larger due to use of data from far larger time differences i e up to 6 hours estimated to be up to 30 K in worst cases over hot land regions in local Summer These errors vary over an image due to the differing geometries of the observations etc but as a guide the average combined error is estimated to be 3 4 K although in certain regions higher values 10 20 K can occur Users of data from the CLAUS archive are strongly advised to appreciate the limitations of the data for example frequency analyses can sometimes produce s
16. paration More recent scientific and technical aspects of the CLAUS project will appear in the CLAUS project area of the ESSC web server Ilear Giida 1 A 1 Cloud Archive User Service 2 The Data This section describes the format of the data files and how they are organised within the CLAUS archive The data are generated and archived at two spatial resolutions e one half degree the original resolution as used in the CLAUS project for climate model comparisons e one third degree started relatively recently these data are being produced simultaneously with version 4 5 of the lower resolution data and are intended for use in process oriented studies and for the creation of monthly animations for educational purposes A brief description of how the data were generated appears in section 3 Fuller details appear in Hodges et al 2000 Information on more recent technical developments and the quality control of the ISCCP data may be found in the CLAUS project area of the ESSC web site 2 1 Introduction The primary data set in the CLAUS archive is a series of three hourly Brightness Temperature Tp images of the Earth stored in the Portable Grey Map PGM format Appendix A 1 This is a simple flat file binary format preceded by an ASCII readable header that contains information such as the image dimensions and version number Figure 1 Sample half degree resolution image from the CLAUS archive for 6 1 19
17. purious results because of the sampling characteristics of the satellites especially the polar orbiters ERBE attempts to account for these effects by modelling the diurnal variation and so correct the data This approach is not used in the CLAUS dataset at present since it might lead to unwarranted confidence in the accuracy of the data However such an approach might be investigated in a future product since the processing required could be applied retrospectively i e to the CLAUS data product rather than having to go back to the source B3 data Ilear Giida 1 4 11 CLAUS 4 Acknowledgements The CLAUS dataset could not have been created without the efforts of many people the ISCCP team for producing and making the B3 data available to the general community EOSDIS in particular the staff at the Langley DAAC for providing a relatively painless way of accessing the large volume of data involved and sorting out various problems and from members of the CLAUS Consortium whose feedback has been invaluable as have comments and suggestions from users worldwide Finally our thanks to Murray Salby and co workers who proved the feasibility and utility of such a satellite derived dataset in climate applications and provided the inspiration for the CLAUS project Ilear Giiida 1 A 19 CLAUS 5 Access to the Archive and Conditions of Use In order to gain access to the data please follow the procedures set out on the BAD
18. r each three hourly Tg image The CLAUS archive is dynamic it is continually being updated using improvements to the processing methodology and quality control of the source data It is being extended to the end of 1999 as further data become available from ISCCP The current status of the archive appears in the CLAUS area of the BADC web site Table 5 outlines the differences amongst the versions By default data are being re processed on a quarterly sometimes monthly basis starting from 1983 However this schedule may be modified by user requests for particular periods so that older versions of the data may exist for some while Recently from version level 4 5 an additional higher resolution product one third degree has starting to be produced These data are also being used to generate monthly animations in Quicktime format which are proving popular for educational purposes Both of these new datasets are being archived at BADC The CLAUS data have been used in a number of scientific studies by the climate modelling groups involved in the CLAUS Project These along with details of the processing methodology are described in the CLAUS Final Report an edited version of this in PDF format may be downloaded from the CLAUS area of the BADC web site CLAUS related publications will appear in various journals over time e g Yang and Slingo 2001 Further papers discussing the technical issues involved in processing the ISCCP data are in pre
19. s The CLAUS archive currently spans the period Ist July 1983 the start of the ISCCP period to 31st December 1995 Responsibility for maintaining and updating the CLAUS archive resides with the NERC Environmental Systems Science Centre Reading University UK Because the data have proved invaluable in validating many aspects of GCMs and are potentially of wider use the CLAUS archive is being made available generally through the British Atmospheric Data Centre BADC subject to terms and conditions of use Section 4 The source data used in creating the CLAUS archive are the level B3 reduced resolution thermal infra red radiances from the operational meteorological satellites participating in the International Satellite Cloud Climatology Project ISCCP and were obtained from the NASA Langley Atmospheric Sciences Data Center LASDC The B3 data consist of several channels channel 1 0 4 um visible channel 2 10 wm thermal Infra red window channel 3 6 3 wm water vapour The CLAUS archive at present is produced from channel 2 which is processed to create a uniform latitude longitude gridded dataset or image of Brightness Temperature Tg values at a spatial resolution of 0 5 by 0 5 degrees and temporal resolution of three hours Hodges et al 2000 Information at the grid point level about which satellites were used in generating each Ts image and the type of interpolation applied is held in two supplementary quality files fo
20. s used to hold the following CLAUS specific information Resolution Either 0 5 or 0 3333 degrees Type Type of data file as text string BT CLAUS Brightness Temperature Image Data CS CLAUS Contributing Satellite Image Data IQ CLAUS Interpolation Quality Image Data Synoptic Date This should correspond to the file name Source Channel ISCCP channel number 10pm TIR in this example Satellites ISCCP satellite numbers see Table 2 corresponding to the bit flags in the es file starting at bit 0 00 denotes that no satellite was operating at that time Creation Date Date and time that the Tg image file was created Revision Version number of the data processing Quality Control Line 9 contains the width and height of image in grid points Line 10 denotes the maximum possible data value 255 Ilear Giida 1 A 1A A2 Monthly Animations The CLAUS monthly animations are stored in Apple s Quicktime format which enables them to be displayed with the same fidelity as the individual brightness temperature images i e without the loss of detail that often occurs with other formats such as MPEG To help differentiate land and ocean regions a bi colour tint currently red for land and blue for ocean is applied to the brightness temperature images before they are processed to form the animations files The date time of each individual image is also written in the bottom right hand corner of the corresponding frame
21. satellites were operational simultaneously plus INSAT 1B More usually however six satellites four geostationary and two polar orbiters were in use the nominal configuration although for certain periods sometimes only four three geostationary and one polar orbiter were available Because the satellite configuration changes over time the Contributing Satellite cs data represented in the form of a PGM image file A 1 are used to record which satellites were used in estimating the corresponding Tg value at each grid point This information consists of a set of eight bit flags arranged in a byte field as shown in Table 1 The bit positions are indices to a set of eight ISCCP satellite codes Table 2 which are contained in the headers of all three PGM files for each three hourly image Figure 5 Ilear Giida 1 A A CLAUS Bit Position Value Nominal Satellite Series 1 rightmost NOAA Afternoon PM NOAA Morning AM GMS O a f 2a Oo n 3 4 4 8 GOES West or GOES Central PCT O 6 32 OoOo 7 64 METEOSAT Indian Ocean INSAT or METEOSAT 5 at 63 E 8 leftmost spare Table 1 Contributing Satellite flag structure The arrangement shown is for the nominal satellite configuration but can vary see text 7 2 4 16 GOES East or METEOSAT 3 at 75 W 32 64 Bit flag 8 is used when the satellite configuration changes usually when a new satellite became operational but data from the exi
22. sting one was still collected e g during change over periods of the NOAA afternoon satellites Note that some satellites changed their nominal position for example METEOSAT 3 eventually replaced GOES 6 when this failed in 1987 For full details of the ISCCP satellite network refer to ISCCP 1987 Designation Longitude NOAA 7 Afternoon PM Morning AM NOAA 12 GOES 6 135W 105W East E Pime oo A Ek Prime GOES E Prime 0 a METEOSAT 5 Prime 63E Table 2 Operational meteorological satellites used in ISCCP The code numbers are used in the PGM image file headers see section 2 6 1 to indicate which satellites are used in generating the Tg data Ilear Giida 1 A K Cloud Archive User Service Figure 3 shows the Contributing Satellite information corresponding to the example image shown in Figure 1 expressed using a colour lookup table Although in principle there could be 256 combinations and therefore 256 different colours there are only about 30 combinations in reality due to constraints on the satellite positions and orbits Figure 3 Satellite Contribution Information corresponding to the data shown in Figure 1 Each colour represents the coverage of a single satellite or where two or more overlap 2 6 2 Interpolation Quality The iq files are used to represent the reliability and nature of the T interpolation employed at each grid point Each byte in the image is split into three parts
23. vert CLAUS data into an ASCII format may be downloaded from the CLAUS area of the BADC site Appendix A describes the flavour of the PGM format used in the CLAUS archive 2 5 Data Scaling and Geographical Arrangement The unsigned byte data values in the brightness temperature files are mapped so that the values b 1 to 255 correspond to Tp 340 to 170 K linearly i e Tp 340 b 1 170 254 The value 0 is reserved to indicate points for which no value exists or for which no value can be temporally interpolated The temperature resolution of the data is thus approximately 0 67 K The one half lo_res and one third hi_res degree resolution data sets are arranged on a uniformly spaced latitude longitude grid with values stored in row order starting at the top left corner of the grid The longitude of the leftmost column corresponds to 0 for both lo_res and hi_res note that the last column is a duplicate of the first The top row of the lo_res dataset corresponds to 89 30 N and that of the hi_res dataset to 89 40 N The lo_res and hi_res image sizes are thus 720 columns by 359 rows and 1080 columns by 839 rows respectively 2 6 Quality Data Files The quality data for both resolutions are stored as two PGM files These contain information on a per grid point basis about the contributing satellites cs and interpolation quality iq respectively 2 6 1 Contributing Satellites Throughout the ISCCP period up to seven
Download Pdf Manuals
Related Search