comparison of geodetic and glaciological mass balance on gulkana
Contents
1. 2 4 DEM Accuracy 2 2 0 2 ee ee A1 Strip Data Dialogue 0 0 020000 022 eee B 1 1974 Aerial Camera Data 0 000 00000000848 B 2 1993 Aerial Camera Data 00000 000800008 8 B 3 1999 Aerial Camera Data 00 000 0000000848 C 1 Cumulative and Net Balance at Index Sites C 2 Area Altitude Distribution 0 0 0 0 0 0 00084 C 3 Control Point Locations 0 0 00 0000 eee C 4 Conventional and Reference Surface Cumulative Balances vil ix Preface This main body of this thesis has been prepared for submission to the Journal of Glaciology Rod March will be the second author on the paper All of the glaciological data was either prepared by him or based measurements made directly by him or others in the USGS Keith Echelmeyer and Will Harrison also spent many hours editing the paper many and helped immensely with ideas for the scientific process but declined to be listed as authors Appendices are included in the thesis that contain important information which did not have a place in the paper The first appendix is a sample work flow for creating digital elevation models DEMs using PCI Geomatics Apex Software Much of the time spent working on the thesis was devoted to learning the software and creating a viable work flow preparing DEMs for differencing Other operators will benefit from following this appendix Also included is a CD so people working with this data2. To figure out the scan direction use image loader to load the first two images of a 26 strip into view 1 with image 1 on the left and image 2 on the right If the overlap region is to left of image 1 the scan direction is left Repeat this for all the strips Enter the correct scan direction for each strip Up and Down do not work and will corrupt you files In the main triangulation window go to reset gt support file gt backup support This will back up your work The data can be restored under the same menu I have had problems with the backup getting corrupted so also back up the entire data directory to another folder by cutting and pasting in windows explorer Exterior Initialize Exterior initialization is the next step Click the initialize solve tab at the bottom of the triangulation window and select exterior initialize You just backed up the support files so click through the first message Run the orientation Now save the triangulation file and exit Open the load imagery dialog load image pairs e g 1_1 and 1_2 one at a time and examine the images to make sure the overlap area is approximately correct If it is not go to image enhancement pairwise rectify If this doesn t make the images line up there is an error probably with the scan direction The program will let you proceed but do not until the images are correct Also do not run exterior initialize after blunder detect and solve or simultaneous solv
3. 1 1 1 5 00 Lp 1150 1650 2150 Elevation m Figure 2 1 Ablation corrections The corrections were tuned to match measured seasonal ablation at index sites represented by vertical dotted lines We also corrected each DEM for total emergence from the photo date to the end of the ablation season over the interval shown in Table 2 2 Any change in the surface elevation of the glacier at a point not due to ablation is due to the emergence velocity it is the vertical component of velocity corrected for the downstream movement of ice Paterson 1994 We define the total emergence to be the cumulative surface elevation change from emergence velocity over the interval It should be noted emergence does not affect glacier wide balance because it is merely a redistribution of mass along the entire glacier Nevertheless we corrected the DEMs for emergence velocity for two reasons 1 to more accurately compare individual DEM points with optical and laser profiles see section 2 1 4 and 2 to more accurately represent the thinning at specific areas at the end of the ablation season A curve was fit through the total emergence measured over the interval at the index sites and adjusted to the shape of the extended mass balance curve see section 2 2 Flow was not measured in 1974 so the average emergence from 16 years data at each index site was used The shape of the curve from 1993 differs from the others because the measured emergence at the mid gl
4. 0 70 75 88 17 66 24 26 1991 3 42 0 18 0 97 0 07 79 30 17 84 25 23 1992 2 69 0 52 0 61 0 24 81 99 18 36 25 84 1993 4 52 2 54 0 47 1 68 86 51 20 90 25 37 1994 3 96 0 87 0 53 0 60 90 47 21 77 25 90 1995 3 29 1 24 0 28 0 72 93 76 23 01 26 18 1996 4 05 0 82 0 61 0 54 97 81 23 83 26 79 1997 4 99 2 43 0 43 1 71 102 80 26 26 26 37 1998 3 49 0 95 0 33 0 66 106 29 27 21 26 70 1999 4 35 1 35 0 07 1 14 110 64 28 56 26 63 38 e e e e e e e e e e e e e e e e e e e e y y y y y y y y y y y 00 BOONNNNNNNNNNA OO 66 OSSOS Sci II O LO e NOCOCO NOOOD OO OM M cODLOLO LONA OO OO DDBDBLLLLLLIDOD DDD DDD DDD NANA NA 006 0084 008 002 w aBuey uoneAs a s 19 9u10 11 a1enbs ul UONNQLISIA SPNWHIV Pay 1919819 Buey no 1967 1974 1993 1999 and interpolating the intervening years The 1967 AAD is from the Table C 2 Area altitude distribution These were as calculated by measuring the AAD in start of the glacier monitoring and the method used to create it is unknown 39 Table C 3 Control point locations The points were surveyed to 0 10 m with GPS relative to the NAD83 horizontal datum and the NGVD29 vertical datum These were used to control the DEMs Site Easting m Northing m Elevation m Croakley 575941 34 7019451 27 2238 08 IGY 576371 37 7017131 45 2001 63 Pewe 576816 78 7012770 41 1152 70 Downdraft 577688 30 701
5. 1 1 if current strip is 1 Reference Image ID 1 Number of images in strip minus 1 X along strip 0 0 Y across strip 160 160 First strip enter 0 0 A 9 Triangulation This is the most difficult and time consuming part of the process It is also the most likely to corrupt files so back up often After every step back up to a separate folder as outlined before Open the triangulation window by clicking preparation triangulation If working on a surface that has changed though time such as a glacier control all the images together using bedrock and control points to tie the images When extracting DTMs only use images from the a single year Setup Setup is the first step This is where you tell the program where the images are in relation to one another and what algorithms to use for triangulation For the most part this is very straight forward if the standard naming convention was used for the images All of the software defaults will work well With the naming convention used the first number is the strip and the second is the image in that strip This window will be fairly straight forward to fill in with the exception of the the strip data information When all the rest of the fields are filled in click the strip data button on the lower right hand of the window This brings up a dialog which is not well explained in the manual The numbers to write in each box are shown above in Table A 1
6. C 1 Cumulative and net balances at index sites Net Balance m weq 37 The index sites balances were measured and calculated by the USGS and the conventional glacier wide balances were calculated for this thesis Cumlative Balance m weq Glacier Year Site A Site B Site D Wide Site A Site B Site D 1966 2 80 0 00 0 88 0 16 2 80 0 00 0 88 1967 2 60 0 50 1 49 0 03 5 40 0 50 2 37 1968 2 73 1 12 1 37 0 16 8 13 1 62 3 74 1969 3 25 1 60 0 23 0 99 11 38 3 22 3 97 1970 2 35 0 05 1 58 0 39 13 73 3 17 5 55 1971 1 85 0 40 1 46 0 28 15 58 3 57 7 01 1972 2 70 1 00 0 88 0 36 18 28 4 57 7 89 1973 1 95 0 11 1 63 0 54 20 23 4 46 9 52 1974 3 75 1 65 0 13 1 12 23 98 6 11 9 65 1975 2 75 0 55 0 84 0 25 26 73 6 66 10 49 1976 3 85 1 50 0 24 0 96 30 58 8 16 10 73 1977 3 26 0 87 0 98 0 24 33 84 9 03 11 71 1978 3 16 0 70 0 92 0 22 37 00 9 73 12 63 1979 3 50 1 26 0 66 0 56 40 50 10 99 13 29 1980 3 10 0 59 1 06 0 09 43 60 11 58 14 35 1981 2 54 0 02 0 88 0 02 46 14 11 60 15 23 1982 3 22 0 67 1 05 0 14 49 36 12 27 16 28 1983 3 30 1 12 1 44 0 00 52 66 13 39 17 72 1984 3 05 0 90 0 74 0 34 55 71 14 29 18 46 1985 2 12 0 26 1 70 0 66 57 83 14 03 20 16 1986 3 14 0 02 1 07 0 04 60 97 14 01 21 23 1987 3 37 0 52 1 02 0 14 64 34 14 53 22 25 1988 3 35 0 87 0 99 0 23 67 69 15 40 23 24 1989 4 25 1 43 0 64 0 71 71 93 16 83 23 88 1990 3 95 0 83 0 38
7. ice Gulkana Glacier has ex perienced much less change than South Cascade Glacier over the measurement periods so any error associated with assuming Sorge s Law will be at most a few percent Seasonal corrections especially in 1993 and relative orientation of the DEMs are the largest errors These are both estimated at 0 3m weq Based on these estimates we take as a conservative 13 estimate of the error in the geodetic balance over each interval to be 0 5 m weq 2 2 Glaciological Balance The USGS has used the glaciological method to determine the net mass balance on Gulkana Glacier every year since 1966 In this method the end of the balance year is defined as the date of the yearly glacier wide minimum balance They maintained an extended stake network of up to 30 mass balance stakes until the mid 1970s when the stake network was reduced to three index sites see Figure 1 2 and measurements were expanded to include ice motion and surface elevation at these sites March 1998 Since 1974 the balance has been calculated using data from these three index sites The highest index site D is generally just above the ELA but the ELA has been above site D three times The seasonal balance is measured in an area 25 75 m around each pole in an effort to reduce errors from individual snow depth soundings and small scale surface irregularities Trabant and March 1999 To calculate the glacier wide balance from the index site measureme
8. individual measurement this demonstrates how accurate a single point can be extracted The standard deviation of the mean is how well the mean offset is known Data Mean Standard Standard Remarks Offset Deviation about Deviation of m weq the Mean mweq the Mean m weq Bedrock 1993 1974 0 20 4 70 0 22 Relative Error Bedrock 1999 1993 0 12 5 20 0 15 Relative Error Optical Survey 1993 DEM 0 74 1 34 0 20 Absolute Error 2000 Profile 1999 DEM 0 21 1 87 0 15 Absolute Error 1993 Profile 1993 DEM 0 49 1 67 0 15 Absolute Error Airborne laser altimetry profiles flown in 1993 and 2000 were measured to an accuracy 12 of about 0 3m Echelmeyer et al 1996 These profiles cover the centerline of the main branches in an almost continuous line down the glacier Sapiano et al 1998 The 1993 and 2000 laser altimetry profiles show the 1993 DEM to be 0 55 m weq low and 0 23 m weq low respectively Figure C 1 These also show no trends in the difference indicating the DEMs are not sloping relative to the datum The mean offset is probably due to the long duration over which seasonal corrections were calculated The important results of the profile comparisons are that the absolute accuracy of the 1993 and 1999 DEMs is less than 1m and that the accuracy of elevations extracted in the accumulation area are satisfactory We did not use these independent profiles as an indication of error among the DEMs We conside
9. 5946 51 1600 03 Yes L 579330 89 7018808 42 1757 82 Blinded 579589 60 7014200 28 1676 42 Pass 580387 03 7019711 15 1909 40 Slim 580403 17 7018237 07 1910 46 Bogus 581389 55 7017035 76 2293 50 Moore 581761 29 7018815 30 2090 36 Table C 4 Conventional and reference surface cumulative balances Year Convetional 1993 1967 Reference Reference Surface Surface m weq m weq m weq 1966 0 16 0 06 0 16 1967 1968 1969 1970 088 0 44 0 99 1971 060 009 074 1972 0 96 040 115 1973 1974 154 090 188 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 12 91 11 65 15 71 Al Bibliography Andreassen L M Comparing traditional mass balance measurements with long term vol ume change extracted from topographical maps A case study of Storbreen Glacier in Jotunheimen Norway for the period 1940 1997 Geografiska Annaler 81A 4 467 1999 Arendt A A Mass Balance Modeling of an Arctic Glacier Masters thesis University of Alberta Edmonton AB 1997 Arendt A A K A Echelmeyer W D Harrison C S Lingle and V B Valentine Rapid wastage of Alaska glaciers and their contribution to rising sea level Science 297 382 2002 Bader H Sorge s law of densification of snow on high polar glaciers Journal of Glaciology 15 319 1954 Braithwaite R J an
10. Apex uses A 2 Terms Minification The largest view without interpolation is 1 1 and each larger scale allows you to see more of the image For each scale e g 1 64 a new image is created during minification so zooming is faster Interior Orientation This process corrects an image for lens distortion Exterior Orientation Triangulation This process orients an image either relative to other images or to true ground coordinates Console Monitor The monitor which displays menus and is not in 3D Extraction Monitor The monitor which displays in 3D Photograph The picture taken by aerial photography Image The digital picture after scanning Fiducial Crosses or dots on aerial photographs which are used to correct for lens distor tion A 3 Scanning Three things are important in scanning the resolution typically measured in pixels per inch ppi the bit depth 10 bit or 8 bit and the image orientation The scan resolution directly affects the accuracy of the software The horizontal accuracy of the software is 1 2 times the ground pixel size the width in ground space of one pixel and the vertical accuracy is 0 5 to 3 times the ground pixel size PCT 2000 The ground pixel size can be 22 computed from the following formula S 1 PM pr 3937 A 1 where PM is the ground pixel size in meters S is the scale and PPI is the scan resolution in ppi Slama 1980 The effectiveness of a higher scan resol
11. B Kennedy Glacier mass balance trends in Alaska and climate regime shifts Eos Transactions American Geophysical Union 79 48 F277 1998 Trabant D C and R S March Mass balance measurements in Alaska Geografiska An naler 81A 4 777 1999
12. COMPARISON OF GEODETIC AND GLACIOLOGICAL MASS BALANCE ON GULKANA GLACIER ALASKA By Leif H Cox RECOMMENDED Advisory Committee Chair Chair Department of Geology and Geophysics APPROVED Dean College of Science Engineering and Mathematics Dean of the Graduate School Date COMPARISON OF GEODETIC AND GLACIOLOGICAL MASS BALANCE ON GULKANA GLACIER ALASKA A THESIS Presented to the Faculty of the University of Alaska Fairbanks in Partial Fulfillment of the Requirements for the Degree of MASTERS OF SCIENCE By Leif H Cox B S Fairbanks Alaska December 2002 iii Abstract The net mass balance on Gulkana Glacier has been measured since 1966 by the glaciologi cal method in which seasonal balances are measured at three index sites and extrapolated over large areas of the glacier Systematic errors accumulate through time in this method therefore the geodetic balance in which errors are independent of time was calculated for comparison to and possible calibration of the glaciological method Digital elevation models DEMs of the glacier in 1974 1993 and 1999 were prepared and geodetic balances com puted giving 6 0 0 5 m of water equivalent weq from 1974 to 1993 and 11 8 0 5m weq from 1974 to 1999 These are to be compared to the glaciological balances over the same in tervals which were 5 84 0 9 m weq and 11 24 1 0 m weq respec
13. Method Accuracy The comparison between geodetic and cumulative glaciological balances is shown in Fig ure 3 1 and Table 2 3 The comparison is excellent with the geodetic balance within the estimated error bars of the glaciological balance This implies that the glaciological bal ance method on this glacier does not have large systematic errors that could arise from several sources including sinking poles erroneous snow depths missing internal ablation and accumulation and an invalid area extrapolation The USGS includes an estimated 0 05m weq a internal ablation in the net balance Systematically ignoring this small fac tor would have decreased the cumulative glaciological balance by about 10 On Alfotbreen Glacier Ostrem and Haakensen 1999 placed plywood at the base of mass balance poles and observed poles forced through the plywood due to snow compaction The USGS circumvented this problem by laying plywood or sawdust on the summer surface to unambiguously locate it the following spring by drilling or coring Trabant and March 1999 In addition single point measurements are not necessarily representative of the immediate area deviations of 0 23 mweq in one year have been observed on three stakes less than 5m apart Braithwaite and Olesen 1989 Errors from these variations can sometimes be eliminated by sampling the balance in an area tens of meters around each index site which the USGS does in both the ablation and accumulation sea
14. Photo Focal Scan Ground Remarks Collected Photographs Scale Length Res Pixel Points mm um Size m Aerial 9 7 1974 4 1 22000 151 293 10 0 22 Missing Minya Photography Basin Monochrome Aerial 7 11 1993 8 1 36000 153 211 10 0 37 Excellent Quality Photography Color Aerial 8 18 1999 9 1 24000 151 830 7 0 17 Poor contrast in upper Photography basins Monochrome Laser Profile 6 12 1993 10 000 Laser Profile 6 3 2000 10 000 Optical Profile 8 1 1993 56 1992 ten control points were surveyed to about 0 1 m using the Global Positioning System GPS Figure 1 1 these were marked on the ground with 10 m x 10 m white crosses that were easily identified on the images acquired the following year Images from 1974 and 1999 lacked these or any other marked control points they were oriented using obvious features such as rock outcrops and mountain peaks When these features are selected in multiple images they become the tie points which allow the uncontrolled images to be aligned with controlled images A total of 170 tie points on bedrock were used as coincident image points for this relative control After the images were properly controlled an algorithm using image correlation auto matically extracted DEMs on 5 m co registered grids which we found was optimal for image correlation To facilitate manual editing which was only needed on an estimated 10 of the glacier the grids were resampled glacier wide to a 25 m spaci
15. acier site B was greater than the index site low in the ablation area Fig 2 2 Thickness changes of ice or snow were converted to water equivalent based on the density of the material lost or gained The small shift in the ELA during the measurement period Figure C 2 and the relatively small balance gradient make the assumption of Sorge s Law 0 5 1974 e I rR 1998 O 1 5 01 qu ae D Doo Bia ee ee E 0 1 LU 5 0 3 ES 1 1 1 0 5 a i 1150 1650 2150 Elevation m Figure 2 2 Total emergence as a function of elevation This was measured over the correc tion intervals at the index sites represented by vertical dotted lines plausible on Gulkana Glacier This law states that the density structure remains constant in an unchanging climate Bader 1954 allowing us to assume that the change in volume is related directly to water equivalent volume by the density of ice 900kgm Paterson 1994 2 1 3 Surface Elevation Change Surface elevation change was calculated by differencing two registered DEMs over the glacier surface Where a surface change varied by more than 5 m from adjacent areas the point was remeasured in the original DEM Points where elevations could not be extracted accurately due to poor contrast were removed and the elevation change interpolated from adjacent locations The intervals 1974 to 1993 and 1993 to 1999 were differenced and the geodetic bala
16. ages are controlled exit triangulation and click extraction gt terrain automatic DTM extraction The setup is explained well in the manual and the default values work well so I will not belabor the process here A few hints on automatic extraction Use a spacing that results in approximately 10 000 points for the first extraction which is small enough to run quickly and large enough to give meaningful results Check it in the 28 interactive edit which is explained below The terrain may need to be broken up into several different areas to be correctly extracted Likely breaks between DTMs are large changes in slope or brightness e g from the glacier to the valley walls You can open several automatic extraction windows at the same time and run them overnight I ve found the best spacing to be 3 5 m Smaller spacing does not increase accuracy and larger spacing seems to deteriorate accuracy Some regions may not have enough contrast to be accurately picked by either you or the computer so it may be beneficial to use triangulated irregular networks TIN The computer will effectively only pick points within a certain confidence interval when using the TIN method with heavy mass point thinning selected This eliminates trying to edit thousands of inaccurate points in a low contrast area A 11 Quality Control and Manual Editing Open Extraction Interactive edit and load a DTM View each automatically extracted grid where large a
17. al corrections Density corrections convert snow or ice volume to water equivalent volume change Table 2 2 Seasonal Correction Interval The duration of seasonal correction is shown with the glacier wide ablation correction Emergence corrections were applied over the same interval Photography End of Interval Glacier Wide Date Ablation Season days Ablation Correction m weq 9 8 1974 9 20 1974 12 0 2 7 11 1993 9 8 1993 59 1 8 8 18 1999 9 26 1999 39 0 3 The ablation corrections were done following the concepts of Reeh 1991 using the simple degree day model of Arendt 1997 Measured summer precipitation temperature and the DEM for each year were input into the model The model was then tuned by varying temperature lapse rates and degree day factors to force the modeled summer balance to match the summer balance measured at each index site Finally the ablation in meters of water equivalent from the date of photography to the end of the balance year was calculated as a function of elevation Figure 2 1 The 1993 corrections are relatively large especially at low elevations because of the long time interval and the fact that this period extended over the most intensive part of the ablation season Internal ablation and accumulation are assumed to be negligible over the intervals 0 00 o 3 E 1 00 lt o 3 2 00 o al 1974 1993 1999 4 00 lt
18. ameters Field Angle Do Degrees um 75 4 15 9 22 5 2 27 5 5 30 6 32 5 8 35 9 37 5 4 40 2 42 5 0 45 24 Principle Points x mm y mm Indicated principle point corner fiducials Indicated principle point midside fiducials Principle point of autocollimation 0 005 0 001 Calibrated principle point point of symetry 0 013 0 015 Fiducial Locations x mm y mm 120 350 0 000 117 503 0 000 0 011 117 804 0 010 117 591 117 409 76 108 117 681 75 217 76 181 117 516 75 119 117 914 Data Strip QQenvdow gt y 32 Table B 2 1993 Aerial Camera Data This camera was used by AeroMap US Anchorage Alaska to photograph Gulkana Glacier in 1993 Camera and Lens Camera Type Zeiss RMK A 15 23 Lens Type Zeiss Pleogon A Camera Serial Numer 111683 Lens Serial Number 112649 Calibrated Focal Length mm 153 211 Distortion Parameters Field Angle D Degrees um 7 5 2 15 3 22 7 2 30 4 35 2 40 2 Principle Points Indicated principle point corner fiducials Indicated principle point midside fiducials Principle point of autocollimation Calibrated principle point point of symetry Fiducial Locations 3 7 2 x mm y mm 1 103 932 103 948 2 103 937 103 914 Bl 6 3 103 894 103 902 E 4 103 941 103 948 a 5 113 010 0 017 6 112 981 0 020 7 0 020 112 981 8 0 010 112 992 33 Ta
19. and glaciological balances Interval Geodetic Glaciological Balance m weq Balance m weq 1974 1993 6 0 0 5 5 8 0 9 1974 1999 11 8 0 5 11 2 1 0 2 1 4 Errors Errors in the geodetic balance can result from seasonal corrections image control and elevation extraction especially in the low contrast accumulation areas Both absolute control of the DEMs to the map datum and relative control between different DEMs account for errors in image control These are addressed separately in the following three paragraphs Fifty six points were surveyed on the glacier in 1993 to an accuracy of 0 1 m Table 2 1 These points were well distributed over the glacier with one profile up each of the main branches and several longitudinal profiles Each of the profiles was subjected to the same seasonal corrections as the DEMs Comparison of DEM elevations to the optically surveyed points shows that the DEM is systematically 0 74 m weq too low Table 2 4 There appears to be no systematic trend to the offset with elevation although the standard deviation is greater for points in the accumulation area The systematic difference was greater than expected and we cannot find a satisfactory explanation for it Table 2 4 DEM accuracy The standard deviation about the mean is greater over the bedrock than ice and the relative error among DEMs is small The standard deviation about the mean shows the accuracy of an
20. aoaaa aaa ee 20 A2 TECOS a tek aa BE ee Be RA EA a 21 AGS SCANNING a pot Rk Paes Se Serr A Bos Pilea at 21 Aj Project Creation asus apa sm o BR Re ee Ooh pee eho a ER a 23 A 5 Camera Calibration A 23 A6 Importing ani S a oe ne A AAA ee 23 At Minitication ss sus a a AA A A a ee i 24 A 8 Interior Orientation 2 0 2 ee 24 Aad Eriangilatione sa so Hew fe egg RSE GP ES Oe Se ee ee at 25 A 10 Automatic Extraction ee 27 A 11 Quality Control and Manual Editing 28 ACID DTM Merge ci pt A PB Ae ee a 28 A 13 Exporting DTMs a a es 28 B Camera Calibrations 30 C Archived Data 34 Bibliography 41 CD of Archived Data Back Pocket List of Figures 1 1 Location Map stais raios A et SR A Bien dig oe AUS oe ets 1 2 Index Site Weather Station and Control Point Locations 2 1 Ablation Corrections x act ee eRe Pw q SA Be 2 2 Total Emergence as a Function of Elevation 2 3 Suface Elevation Change in Meters of Water Equivalent 2 4 Glaciolovical Balances caia rr MS ee GOR HA Gk SEEN 3 1 Comparison of Cumulative Glaciological and Geodetic Mass Balances C 1 Laser Altimetry Comparison C 2 ELA Variations Through Time o e List of Tables 2 1 Data Collected on Gulkana Glacier 000 2 2 Seasonal Correction Interval 0 0 2 0000848 2 3 Cumulative Geodetic and Glaciological Balances
21. arge effect on sea level Houghton et al 2001 Arendt et al 2002 However world wide there are only 33 glaciers that have a balance record over 40 years in length Dyurgerov and Meier 1997 so one or two glaciers are often used to represent hundreds of glaciers in a region e g Meier 1984 While region wide extrapolations may cause inaccuracies it has been shown that in some areas a single glacier can represent the mass balance of a region Rabus and Echelmeyer 1998 A more fundamental problem is the accuracy of the limited number of mass balance records used for extrapolation The conventional method to measure mass balance which we refer to as the glaciological method relies on balance measurements made at a number of discrete points These are then extrapolated over the glacier usually based on the area altitude distribution AAD strem and Brugman 1991 Accumulation of errors can be problematic in this method We are concerned primarily with systematic errors because they increase linearly with the the number of years N in the record whereas random errors increase as VN It is thus important to check whether the errors in the glaciological method are predominately random or if a large systematic component is present in the given balance record An independent method used to check and possibly calibrate the cumulative glaciological balance is the geodetic method Fountain et al 1997 In this method maps of a glacier or th
22. ble B 3 1999 Aerial Camera Data This camera was used by the Bureau of Land Management Anchorage Alaska to photograph Gulkana Glacier in 1999 Camera and Lens Camera Type Wild RC8 Lens Type Wild Universal Aviogon Camera Serial Numer 485 Lens Serial Number UAg 263 Calibrated Focal Length mm 151 83 Distortion Parameters Field Angle D Degrees um 7 5 5 15 7 22 7 6 30 0 35 6 40 7 Principle Points x mm y mm Indicated principle point corner fiducials 0 004 0 011 Indicated principle point midside fiducials 0 015 0 008 Principle point of autocollimation 0 000 0 000 Calibrated principle point point of symetry 0 007 0 007 Fiducial Locations 3 7 2 x mm y mm 1 106 002 105 984 d 2 106 000 106 110 a 5 6 3 106 010 106 140 s 4 105 992 105 980 a 5 110 006 0 010 6 109 994 0 005 7 0 015 110 001 8 0 014 109 984 Appendix C Archived Data 34 3000 3500 4000 4500 5000 93 DEM v Profiles 1500 2000 2500 Distance along Profiile m 1000 500 w a9ua1ayIg Figure C 1 Laser altimetry comparison The differences between the seasonally corrected 1993 laser profiles and the 1993 DEM show no systematic trends with elevation 36 000 S661 0661 9864 J29A 0861 9464 0261 S961 009 L 069 002 F 062 008 L 088 0061 0s6 L w uogens a Figure C 2 ELA variations through time Table
23. d O B Olesen Detection of climate signal by inter stake correlations of annual ablation data Qaman rss p Sermia West Greenland Journal of Glaciology 35 120 253 1989 Conway H L A Rasmussen and H P Marshall Annual mass balance of Blue Glacier USA 1955 97 Geografiska Annaler 81A 4 509 1999 Dyurgerov M B and M F Meier Year to year fluctuations of global mass balance of small glaciers and their contribution to sea level changes Arctic and Alpine Research 29 4 392 1997 Dyurgerov M B and M F Meier Twentieth century climate change evidence from small glaciers PNAS 97 4 1406 2000 42 Echelmeyer K A W D Harrison C F Larsen J Sapiano J E Mitchell J Demallie B Rabus G Adalgeirsdottir and L Sombardier Airborne surface profiling of glaciers a case study in Alaska Journal of Glaciology 42 142 538 1996 Elsberg D H W D Harrison K A Echelmeyer and R M Krimmel Quantifying the effects of climate and surface change on glacier mass balance Journal of Glaciology 4159 649 2001 Fountain A G and A Vecchia How many stakes are required to measure the mass balance of a glacier Geografiska Annaler 81A 4 563 1999 Fountain A G R M Krimmel and D C Trabant A strategy for monitoring glaciers U S Geological Survey Circular 1132 U S Geological Survey Denver CO 1997 Haakensen M Glacier mapping to confirm results from mass bala
24. ding more detail when needed The following will be enough to get one started and reproduce the DEMs used for this thesis This is by no means a comprehensive manual The scanning section should be read prior to scanning but the rest of the sections will probably make little sense without the program running in front of you Apex is very finicky it often will corrupt files and then save them upon exiting I recommend backing up the entire data directory every time much progress has been made by copying the entire file into anther directory using windows explorer Commands are also often grayed sometimes because a window was closed with the x instead of file gt exit or sometimes it just seems to happen This is fixed by exiting and reopening the program Apex saves all progress automatically on exiting so if an error is made before exiting the data will have to be reloaded from a backed up version There are also shortcut keys for virtually every Apex command familiarization with these will speed repetitive processes Apex is well suited to creating accurate digital terrain models DTMs but is not effective for DTM analysis Another program such as AutoCad with the Quicksurf addition 21 or ARC should be used for anything but the most basic analysis DTM is interchangeable with digital elevation model DEM as used in the paper I used DEM in the paper because it is more common in literature but use DTM here because this is the terminology
25. e 258 5079 115 1992 strem G and M Brugman Glacier mass balance measurements Science Report 1 Na tional Hydrology Research Institute 1991 strem G and N Haakensen Map comparison or traditional mass balance measurements Which method is better Geografiska Annaler 81A 4 703 1999 Paterson W S B The Physics of Glaciers third ed Pergamon Oxford 1994 PCI Apex version 7 0 user s manual PCI Geomatics Richmond Hill Ontario Canada 2000 P w T and L Mayo Guidebook to permafrost and Quaternary geology along the Richard son and Glen Highways between Fairbanks and Anchorage Fourth international confer ence on on permafrost vol 1 chap Delta River Area Alaska Range p 47 Alaska Division of Geological and Geophysical Surveys 1983 Rabus B T and K A Echelmeyer The mass balance of McCall Glacier Brooks Range Alaska U S A its regional relevance and implications for climate change in the Arctic Journal of Glaciology 44 147 333 1998 Reeh N Parameterization of melt rate and surface temperature on the Greenland Ice Sheet Polarforschung 59 3 113 1991 Sapiano J J W Harrison and K Echelmeyer Elevation volume change and terminus retreat changes of nine glacier in North America Journal of Glaciology 44 146 119 1998 44 Slama C C Manual of Photogrammetry American Society of Photogrammetry Falls Church VA 1980 Trabant D R March and
26. e This can cause file corruption Control Point Editor Bring up the control point editor by clicking preparation control point editor Now select File Select GPF and name a new file Enter the control point names and coordinates making sure to click accept before adding a new point Include the accuracy because it will be used in the final triangulation solution Save this and exit Interactive Point Measurement Go back into triangulation and click on interactive point measurement IPM If many of the commands are grayed out select view 0 in the display utility In the IPM window click grnd file and select the file that you just created in the control point editor The IPM process is fairly straight forward and well described in the manual Pick at least four points control or tie in each image and make sure each strip is tied together with at least four points Save and exit after picking points 27 Blunder Detect and Solve Click on initialize solue blunder detect and solve Dese lect point distribution for now This can be useful for checking the control and tie points distribution later but first you want to get the images initially oriented Click start Clicking on edit failures will tell you which images need more tie points or if there is a bad point You can adjust the image parallax and other parameters to make this less rigorous Iteratively run this tool and interactive point editor until blunder detect and solv
27. e is successful If things are not going well check the image orienta tions as before with the image loader If the images are not close to their expected relative locations i e the overlap region is not correct you will have to reload the backed up support files and add tie points or correct erroneous ones Simultaneous Solve After blunder detect and solve has been successful run simultane ous solve This will perform a least square inversion on all the images simultaneously Click on initialize solue simultaneous solve After hitting start check the image pixel residual the manual recommends this be lower than 1 for a final solution If the whole solution is poor i e the image residual is greater than 5 tie or control points will have to be corrected or tie points added If it is very high gt 1000 there is an erroneous data point or an error in setup Note that the window will not resize to accumulate large numbers so a huge value such as 1 23456789e 120 might be inter preted as 1 23456 If the residual is very large you will need to reload the backed up files and start again at setup Click on display residuals This displays all control and tie points in ascending order of accuracy Points with large residuals will have to be remeasured This can be done either from the current window or in the interactive point editor Keep adjusting points until an acceptable solution is reached A 10 Automatic Extraction Once the im
28. ead to such patterns of change 18 Chapter 4 Conclusions The agreement of the two mass balance methods on Gulkana Glacier is encouraging It supports the use of the limited number of index sites for determining the net glaciologi cal balance if sufficient care is make in the required measurements and corrections The glaciological mass balance of Gulkana Glacier can be accurately represented by three index sites with only one accumulation area site located just above the average ELA This also demonstrates that the balance in a small radius can accurately describe the balance in an elevation band and the extrapolation with elevation and area has no large systematic errors on this glacier This does not necessarily apply to other glaciers or even to future measurements of Gulkana Glacier Every glaciological mass balance record needs to be regularly calibrated with the geodetic method The relatively small error of the carefully measured glaciological balances makes them ideal for annual measurements and the time independent nature of the geodetic method makes it ideal for long term several years to decades measurements Many glaciers have featureless accumulation areas this can account for large errors in the geodetic balances The accumulation area on Gulkana Glacier is broken up into several small cirques with numerous nunataks and crevasses that aid stereo viewing The snow line was also anomalously high during each year of photography p
29. ecially when relating bal ance to climate Elsberg et al 2001 but must be stated as such when published because they are not the actual mass change of a glacier and cannot be directly compared to the geodetic balance and hydrologic outflow For accurate comparisons the conventional glacio logical balance as shown in Figure 3 1 and Table 2 3 is needed The geodetic balance can be used to correct the reference surface balance to a conventional balance using a one or two parameter fit as outlined by Elsberg et al 2001 17 3 2 Climate and Glacier Response The geodetic balances correspond to an average annual thinning rate of 0 31ma from 1974 to 1993 and 0 96 ma from 1993 to 1999 This accelerated thinning rate in the 1990s has been observed nearly everywhere in Alaska Arendt et al 2002 The more continuous cumulative glaciological balance record shows these trends as well Figure 2 4 The total thinning is much greater near the terminus over the first period Figure 2 3 From 1974 to 1993 we found a maximum thinning of 60m weg compared to the second interval with a maximum of 40mweq However during the first interval there is little change in the accumulation area while in the second interval there was 4m weq thinning in the accumulation area This trend has been witnessed in most of Alaska Arendt et al 2002 Lower net accumulation rates in the accumulation area accompanied by a general glacier velocity decrease would l
30. eir digital equivalent digital elevation models DEMs are created using photogrammetry at intervals of a few years to a few decades Differencing these DEMs after applying corrections for density and other factors gives the glacier wide cumulative balance for the different time intervals This method accounts for all spatial variability in balance assuming the DEMs are accurate everywhere and references a stable geographic datum These two methods attempt to measure the same quantity the glacier wide balance but the results differ because of errors inherent in each method Previous studies comparing the results of the two methods have found systematic errors in the glaciological method and map errors in the geodetic method Errors in the glaciological balance have resulted from poles sinking into the snowpack incorrectly defined previous seasonal surfaces and or un accounted for internal accumulation Krimmel 1999 Haakensen 1986 Conway et al 1999 These errors will be summed over the glacier and combined with cross glacier variations in balance from surface irregularities avalanches wind deposits or scours and topographic shading Fountain and Vecchia 1999 Krimmel 1999 Such errors can accumulate system atically over time which has been shown even to cause the glaciological balance to have the opposite sign of the geodetic balance Conway et al 1999 Errors associated with the geodetic method are primarily due to poor photogramm
31. etric contrast in high accumulation areas and poor DEM registration which in extreme cases can cause the balance to vary by several times the accepted value strem and Haakensen 1999 Andreassen 1999 These errors do not accumulate over time making the cumulative geodetic balance more accurate than the cumulative glaciological balance over time scales longer than a few years The geodetic method can thus be used to calibrate the glaciological cumulative balance Elsberg et al 2001 The mass balance record is especially important on Gulkana Glacier because it has one of the longest mass balance records in the United States 1966 present It is one of three index glaciers chosen for long term balance monitoring by the United States Geological Survey USGS and is the only one of these index glaciers in a continental climate zone It is often used for studies in glacier climate interaction and sea level change Letr guilly and Reynaud 1989 Trabant et al 1998 Dyurgerov and Meier 1997 The accuracy of the glaciological balance on Gulkana Glacier has not been independently verified Here we determine the geodetic balance over two intervals for comparison to and possible calibration of the glaciological record o ee ma Fairbanks y ae Gulkana Glacier o Anghorage Edge E Be ea die a En o Kilometers E Pia e PO O 250 500 1 000 Figure 1 1 Location map Gulkana Glacier is located in the eastern A
32. has an M S and claims to not need another Many of the other members of the lab both students and faculty read a version of the thesis at least once if not many times Carl Benson Adam Bucki Dan Elsberg and Sandra Zirnheld Thank you I didn t give By Valentine a change to read the manuscript but I didn t want to ruin her perpetual cheerfulness I also did not give Craig Lingle a change to read it but if it wasn t for him I might never have become comfortable with the kinematic surface boundary equation My wife Trilby Cox provided most of the support and funding for the non work parts of my tenure as a masters student Thank you so much especially for your support the last two months of thesis writing which we both thought might never end I would also like to thank my parents both sets and my little sister Heidi for late night e mails and other sources of encouragement This project was funded by the United States Geological Survey Many thanks to Dennis Trabant and Rod March for securing funding and anyone else in the USGS who helped Chapter 1 Introduction Glacier wide net mass balance is the net gain or loss of mass over an entire glacier during a given balance year summing the net balance over a series of years results in a cumulative balance Regional trends in cumulative mass balance are indicators of climate variability Oerlemans and Fortuin 1992 Hodge et al 1998 Dyurgerov and Meier 2000 and can have a l
33. ing aerial photography from 1974 1993 and 1999 Table 2 1 See Appendix B for camera calibrations and photography credits AeroMap US Anchorage Alaska had previously created a high quality DEM from the 1993 photographs but due to problems transferring control points to the 1974 and 1999 images and for increased relative accuracy we made another DEM from the 1993 photographs Photos taken in 1974 are high quality but lack coverage in the upper Minya Basin as will be discussed below 2 1 1 DEM Creation We generated three DEMs of Gulkana Glacier from the aerial photographs using a digital photogrammetry system PCI Geomatics Apex The digital process is very similar to analytical photogrammetry except that the photographs are scanned to create a digital image the extraction of elevations is semi automated and a 3 D viewing system is used to edit the DEMs Scan resolution limits accuracy with a horizontal accuracy equal to about 1 to 2 times the ground pixel size the ground dimension represented by one pixel and a vertical accuracy of 0 5 to 3 times the ground pixel size PCT 2000 Two types of control are used to orient images control points which orient the images to absolute ground coordinates and tie points which align the images to each other In Table 2 1 Data collected on Gulkana Glacier assess DEM accuracy This data was used to prepare DEMs and Data Date Number of
34. laska Range 63 16 N 145 25 W Figure 1 1 It has three accumulation cirques facing approximately south east south and west with the maximum elevation of 2450 m in the south east facing cirque known as the Minya Basin Ice from the three accumulation areas merges below the average equilibrium line altitude ELA of 1780m and flows south to the terminus at 1200m above sea level March 1998 The terminus has retreated 3km since its Little Ice Age maximum at the turn of the 20 century P w and Mayo 1983 and about 300m since 1974 Glacier area has decreased from 18 4km in 1974 to 17 1 km in 1999 The average balance gradient is 0 5 m a7 100 m Air temperature and precipitation have been measured since 1967 at a weather station lo cated on a moraine east of the lower glacier the record is 93 complete Kennedy et al 1997 Presently there are three index sites on the glacier labeled as A B and D as shown in Figure 1 2 at which surface motion and mass balance have been measured by the USGS since the mid 1970 s March 1998 Three laser altimetry elevation profiles were flown in 1993 1995 and 1999 and the glacier elevation profile was optically surveyed in 1993 E Weather Station Average ELA e Index Sites 4 Control Points Figure 1 2 Index site weather station and control point locations Chapter 2 Methods and Results 2 1 Geodetic Balance Geodetic balances for Gulkana Glacier were calculated us
35. mage viewing software to make sure they are readable A 7 Minification Go to the main window again and select preparation minification Select an image you want to minify it but don t click start Open another minification window and select a different image Do this with all the images then start all They will take about 15 minutes each When this is finished again load the images and make sure they are visible this time you should be able to zoom out Anytime you make a change to an image you will have to close and reopen the load image window for it to load the updated image A 8 Interior Orientation Use the manual interior orientation under preparation interior orientation manual inte rior orientation Whenever picking specific points on an image use the extraction monitor Sometimes the console monitor does not register points correctly Locate the first two fidu cials accepting each one and then click locate all fiducials Check the locations that were automatically located You can either move the point or click accept and move on The residual should be less than 1 0 according to the manual but I haven t had luck getting the residuals below 2 Make sure you save before loading the next image 25 Table A 1 Strip data dialogue This is a guide for entering data in the table Current Image ID 1 The number of images in the current strip Ref Strip ID Current strip number Same as previous box minus
36. nce from 1974 to 1999 was simply the sum of the two intervals Any errors in the 1993 DEM will be removed in this summation so no inaccuracy was included The upper 2 7km of the Minya Basin did not contain registered grids to subtract be cause the 1974 photography did not cover this area and the 1999 photographs had low contrast Thus the 1974 to 1993 surface elevation change in the Minya Basin was extrap olated from surrounding regions It was assumed to have no surface change because 1 the high areas of the glacier for which coverage existed showed no change over the interval and 10 2 the surface change of the lower Minya Basin trended to zero at the edge of coverage The 1999 TIN in the upper Minya Basin was subtracted from coincident 1993 grid points and the surface change was interpolated between measured points When all the corrections were applied the elevation changes shown in Figure 2 3 were obtained The cumulative geodetic balance then is equal to the glacier wide average surface elevation change integrated over Figure 2 3 For 1974 to 1993 the geodetic balance was 6 0 0 5 m weq and it was 5 8 m weq from 1993 to 1999 Addition of these two balances leads to a strongly negative balance over the entire interval from 1974 to 1999 as shown in Table 2 3 Figure 2 3 Surface elevation change in meters of water equivalent The maps are the two intervals 1974 1993 a and 1993 1999 b 11 Table 2 3 Cumulative geodetic
37. nce measurement Annals of Glaciology 8 73 1986 Hodge S M D C Trabant R M Krimmel T A Heinrichs R S March and E G Josberger Climate variations and changes in mass of three glaciers in western North America Journal of Climate 11 9 2161 1998 Houghton J T Y Ding D Griggs M Noguer P van der Linden X Dai K Maskell and C Johnson Climate Change 2001 The Scientific Basis Cambridge University Press New York 2001 Kennedy B W L R Mayo D C Trabant and R S March Air temperature and precip itation data Gulkana Glacier Alaska 1968 96 Water Resources Investigations Report 97 358 U S Geological Survey Fairbanks AK 1997 Krimmel R M Analysis of difference between direct and geodetic mass balance measure ments at South Casade Glacier Washington Geografiska Annaler 81A 4 653 1999 Letr guilly A and L Reynaud Spatial patterns of mass balance fluctuations of North American glaciers Journal of Glaciology 35 120 163 1989 43 March R S Mass balance meteorological ice motion surface altitude and runoff data at Gulkana Glacier Alaska 1994 balance year Water Resources Investigations Report 97 4251 U S Geological Survey Fairbanks AK 1998 Meier M F Contribution of small glaciers to global sea level Science 226 4681 1418 1984 Oerlemans J and J P F Fortuin Sensitivity of glaciers and small ice caps to greenhouse warming Scienc
38. ng Andreassen 1999 has demonstrated this is a suitable grid spacing for geodetic balance calculation Manual editing was needed in areas of low contrast such as the upper Minya Basin where bright snowfields display few features to be correlated Manual editing is difficult in these areas and care must be taken to not float the grid points in bright areas to a higher elevation than dark areas for lack of other information Grids were not extracted from the 1999 Minya Basin because of poor contrast instead a triangulated irregular network TIN was used Unlike the grid method in which the software picks a point at every grid node regardless of accuracy the TIN method effectively only extracts points with an image correlation coefficient greater than a specific value This eliminates the process of the operator manually editing thousands of inaccurate points although some accuracy is lost because the point density may be reduced as much as an estimated ten times in very poor contrast areas 2 1 2 Corrections Before differencing there are three corrections that we need to apply to each DEM to get geodetic balances that can be directly compared to the glaciological balances ablation emergence and density Ablation and emergence corrections were applied because the date of photography did not coincide with the end of the ablation season which is when the elaciological balance is measured see Table 2 2 We refer to these as season
39. nts the glacier is divided into three elevation bins each centered at elevations halfway between the elevations of sites A and B and sites B and D The elevation bins are updated each year for the variations in index site altitude but not for changes in glacier area The map area of each elevation bin is divided by the total area of the glacier to obtain an area weighting factor This is equivalent to using an area altitude distribution The weighting factor is multiplied by the balance at each site and the results are summed to determine the glacier wide surface elevation change Estimated internal ablation from geothermal heat ice motion and water flowing through and under the glacier is added to the surface balance to find a glacier wide net balance March 1998 The 1967 area altitude distribution AAD was used to calculate the area weighting factors from 1966 to 1993 A new AAD was recalculated in 1993 by the USGS and all subsequent balances have used the 1993 AAD For consistency all prior balances starting with 1966 were recalculated by the USGS with the 1993 AAD All previously published balance measurements are therefore referenced to a fixed AAD which effectively yields the reference surface balance of Elsberg et al 2001 This balance is the appropriate variable to compare to climate but it does not represent the true volume change and cannot be directly 14 compared to the geodetic balance For this comparison we have es
40. r geoset images MyProjectName folder select this location If they are stored elsewhere create a new location as before with a new name for the location and the path directing the program to the image location 2 From the file menu select the correct camera calibration for the images 3 Select file import other gt image Browse for the image you would like to import and select it It should be in the folder you specified in location 4 This will bring up the photo data window the coordinates refer to the location of the image corners in millimeters relative to the image center Typically for an image with the data strip on the left the upper left will be x 114 y 114 and the lower right will be 114 114 This is rather unimportant as interior orientation will correct any problems with this 5 On the create files tab select support only 24 6 Go to options and deselect the auto minify option Click start and the image should be imported 7 Repeat this for all images making sure to apply the right camera calibration if using images from different cameras and change the location as needed Make sure the images have loaded properly by exiting from import and selecting in the main window file load images select view 1 and load the images If nothing comes up click move to load point in the display utility window You will not be able to zoom out at this point If you see nothing check the images in any kind of i
41. re is room on the drive where Apex is located transfer the images to the folder you just created If there is not room on the local hard drive the images can be put anywhere on the network Renaming the images to a standard convention will simplify the triangulation process Name the first image from strip one 1_1 and the second image in the same strip name 1_2 etc A strip is a series of images taken on one flight line and the image number is the order the images where taken on that flight line The numbering order of the strips does not matter and images can be labeled in the reverse order in which they were taken 23 A 4 Project Creation Open project creation with preparation create edit project Click on location new loca tion type the project name in the first column and the path to folder you created in the second then save it Create the project as outline in the manual A 5 Camera Calibration The camera calibration is straight forward and described well in the manual Camera calibrations are kept on record for a long time so they can usually be found through the organization that originally took the photos Camera calibration used for this project can be found in Appendix B A 6 Importing Once the camera calibration is entered import the images into Apex Select preparation import image frame to bring up the correct window 1 Click on the location box If the images are stored in the apex_v70 us
42. reas are inaccurate either change the grid spacing to a smaller grid or break the area into more regions You should not have to manually edit large parts of the DEM Rerun inaccurate regions in automatic extraction Once all the grids are acceptable with only small amounts of manual editing needed use DTM merge to resample the grid to a larger grid size I ve found 25 m spacing to be accurate but not an impossible number to manually edit this will depend on the size of the area and the accuracy needed A 12 DTM Merge Select extraction merge DTM merge will resample grids to larger or smaller spacings change the area covered by a DTM and combine multiple DTMs It is explained well in the manual A 13 Exporting DTMs This is a fairly straight forward process outlined well in the manual Use the xyz format to export the DTMs for evaluation in another program Because Apex puts a header on 29 the exported ascii file you may need to open the file in a text editor and erase the heading before using it in another program Appendix B Camera Calibrations 31 Table B 1 1974 Aerial Camera Data This camera was used by Austin Post of the USGS to photograph Gulkana Glacier in 1974 The photographs are located in the ICA collection Geodata Center UAF Fairbanks Alaska Camera and Lens Camera Type Lens Type Camera Serial Numer KC 1B 67 208 Lens Serial Number 475 Calibrated Focal Length mm 151 283 Distortion Par
43. red relative error to be a much better indication of this The relative accuracies of the DEMs were checked by subtracting two DEMs over bedrock The relative error among DEMs is more important than the absolute error of the DEMs when calculating geodetic mass balance because it indicates the total systematic error in the geodetic balance We note that there are several factors which may make the point measurements over bedrock less accurate than those over ice The first problem is that the photographs were scanned to optimize contrast over the ice and snow areas making the bedrock dark and often black in many areas Second bedrock areas were not man ually edited as carefully as the glacier areas Third except for a few locations the only snow free areas near the upper glacier are nearly vertical causing large elevation errors from small horizontal registration errors The increased standard deviation about the mean of the DEM over bedrock compared to measurements on the glacier illustrates these problems In spite of the difficulties the bedrock differencing gave encouraging results with less than 0 3 m error among DEMs Our error budget for the geodetic balance includes elevation extraction emergence cor rections density ablation corrections and relative orientation Elsberg et al 2001 showed that geodetic balances on South Cascade Glacier would have changed by less than 5 5 if firn is lost in the mid elevations of the glacier instead of
44. roviding better contrast at high elevations There was excellent relative control due to numerous tie points with concurrently made DEMs Ablation corrections were calculated using a temporally and spatially tuned model 19 We recommend several steps that can produce more accurate comparisons and possible calibrations of the glaciological method The photography should be taken as close to the end of the balance year as possible this will decrease the amount of error due to seasonal corrections and decrease the amount of snow at higher elevations aiding stereo perception Second if possible mass balance poles should be surveyed near the time of aerial photography so the balance between photography and the end of the ablation season can be determined It is useful to have an independent profile of the glacier surface to check the DEM A reference surface balance which is at present published by the USGS cannot be directly compared to actual mass change The conventional balance which includes changes in the total glacier area and AAD is needed for an accurate comparison 20 Appendix A Supplement to the Apex Manual A 1 Before You Start Apex is a digital photogrammetry system written by PCI Geomatics PCI 2000 The published manual for the software lacks an organized work flow and some directions are ambiguous This appendix is designed to go along with the manual outlining the steps necessary to create accurate DEMs and provi
45. son Trabant and March 1999 Even with perfect point balance measurements three poles have not typically been 16 m weg I onr 10 a Conventional 1967 Reference Surface 154 ON Geodetic Cumulative Balance 1975 1980 1985 1990 1995 2000 Year Figure 3 1 Comparison of cumulative glaciological and geodetic mass balances The geode tic is shown to be within the random error gray of the glaciological balance The 1967 reference surface balance is shown to differ from the conventional balance found sufficient to accurately determine the mass balance on glaciers the size of Gulkana a minimum of 5 to 10 poles is recommended strem and Brugman 1991 Fountain et al 1997 Krimmel 1999 found differences of 30 on the relatively small 2km South Cascade Glacier Washington in a 27 year balance record this was in part due to area extrapolation using between 1 and 20 poles Our results seem to indicate that three poles are adequate to represent the balance on Gulkana Glacier This could be due to the balance curve being well correlated across the glacier and with elevation and the pole locations accurately represent the areas intended But it is surprising that one index site located generally just above the ELA represents the entire accumulation area The reference 1967 reference surface balance differs from the actual mass change as shown in Figure 3 1 Reference surface balances are valuable esp
46. tablished time variable AADs by calculating the area altitude distribution from DEMs of 1974 1993 and 1999 These were then interpolated for the intervening years see Table C 2 The glaciological balances presented in this paper are conventional balances and were calculated using these time variable AADs Again there is a trend toward more negative balances in the 1990s as shown in Table 2 3 and Figure 2 4 The comparison between the conventional and reference surface balances are shown in Figure 3 1 and Table C 4 The published error for the glaciological balance on Gulkana Glacier is 0 20ma March 1998 To verify this the USGS calculated balances for 1966 and 1967 using both the expanded pole network and the three index sites The expanded pole network reduces much of the balance interpolation with elevation and the difference between the two meth ods was within 0 2 ma However the expanded stake network result was not used for calibration nor does the difference indicate whether the error is systematic or random Net Balance o A p a 2 a i l CS 1970 1980 1990 g E Cumulative Balance E o O es 8 10 15 20 1970 1980 1990 Figure 2 4 Glaciological balances a net balance and b cumulative balance The cu mulative balance is bounded by random dark gray and systematic light gray errors of 0 2m weqa t 15 Chapter 3 Discussion 3 1 Glaciological
47. tively both balances show a tripling in thinning rate in the 1990s These cumulative balances differ by less than 6 For this the glaciological method on Gulkana Glacier must be largely free of systematic errors and use a changing area altitude distribution Relatively good contrast in the accumulation area of the glacier increased accuracy in the geodetic method iv Contents Signature Page ora fa ee a ele Bo a A Ba es E a es i Title Parera e Se E LES EE OE Ee A BE bases Ox Ee E ii Abstract g acdsee et yo Rs Pe ee we pied Se a Boe a ea a iii CONTENTS ss dus Dit ORE a E ok des ee SU a a e eee iv List Of EIGUT S vu ica Ht ke Stage Sed o se ee A Dye gs Ht a See a gee vi List of Tables calma ee te eR An ae PE Oa AO A ek a vii List of Appendices a a a A viii Preface dao EA ERA Baie ee AAA AAA PS ber Ae ix 1 Introduction 1 2 Methods and Results 5 2 1 Geodetic Balanc s s o com sa 6 bea ee sm aaa ee ee eae 5 211 DEM Creation sp epa E Sees E E RES A ee ms 5 21 2 COrrections s cus pone esp a Bote ete at fh 2 1 3 Surface Elevation Change 0 2 202 0004 9 QUA A A asas Gs ged heck Spd gs SRU Ra hee Bg EAS Ge Bee kee Beh he SP ee 11 2 2 Glaciological Balance o ee 13 3 Discussion 15 3 1 Glaciological Method Accuracy o 0000 ee eee 15 3 2 Climate and Glacier Response 0 000 ee eee 17 4 Conclusions 18 A Supplement to the Apex Manual 20 A1 Before You Start
48. ution will decrease when the grain of the photographs becomes visible Less than perfect image control and lens distortion will also decrease accuracy Different bit depths can be specified with color typically scanned in 8 bit while 10 bit is reserved for black and white Apex currently does not use the full 10 bits for automatic extraction but the images can be viewed and manually edited in 10 bits This was tried with the black and white photos on Gulkana Glacier and little was gained in terms of increased contrast from 10 bit depth The images required 4 times more memory and do not export well so I would recommend using 8 bit depth The scanning orientation does matter with Apex The images must be either scanned left or right as defined by the Apex manual Left or right scan direction relates to the overlap region of the image If you display the images side by side the overlap must either be on the left or the right not the top or bottom If the photographs are scanned with incorrect orientation it can be corrected in Apex but is time consuming One other consideration is as of this writing Apex cannot handle images over 2 Gb When you transfer the scanned images to the computer create a folder in the Apex directory under apex_v70 usr geoset images called MyProjectName You will need room for all the images plus room for minified images which take up again as much room as the original images plus about 1 Gb for other files If the
49. will not have to reenter it by hand All of the data on the CD is also shown in appendices because it is unknown how long the data format on the CD will be compatible with typical computers or even how long CDs will be used Will Harrison contributed a great amount to this thesis and had many ideas to improve it I learned from him that it s the systematic errors that will kill you and many other things not fit for a thesis He works too hard for a man who is retired but his help was greatly appreciated Rod March had much patience helping me to find archived data and helping with my computer when it would crater every week or so Roman Motyka also read through the manuscript several times and suggested ways to improve it Martin Truffer had the bad luck to be the only professor consistently around so he took the brunt of my questioning during the writing and also read the thesis numerous times I would like to thank Keith Echelmeyer for the time he put in to help me in spite of his adverse situation During my time at UAF I learned an incredible amount about glaciers and climbing from him he also tried to beat geodynamics into me and I think some of it stuck More importantly he taught me about being persistent and careful in my work and even more about life in general Anthony Arendt contributed so much to this paper especially in the initial stages that I was worried he might claim first authorship Luckily he already
Download Pdf Manuals
Related Search