scribed 
vations 
  
  
  
SE by 
  
  
  
  
ncan 
(ED and 
1b/scrub, 
crops, 
Figure 8 
rall true 
t a mean 
ses, so if 
ose open 
or much 
d several 
elevation 
ly two- 
ring the 
»verall 
  
  
  
mean error (bias) in GDEM v2 when compared with 
GDEM v1 (-0.20 m vs. —3.69 m). 
e [tis clear that GDEM v2 includes non-ground-level 
elevations for areas that have aboveground features (tree 
canopies and built structures). Table 2 shows how the 
mean error increases in the developed land cover classes as 
the number and density of built structures increases. This 
condition is observed in both the comparison of GDEM v2 
with GPS benchmarks, which represent ground level 
elevations, as well as in the GDEM v2-NED differencing, 
with NED representing ground level elevations. 
e In many forested areas, GDEM v2 has elevations that are 
higher in the canopy than SRTM. This observation is 
based on both the comparison of GDEM v2 with GPS 
benchmarks, as well as the GDEM v2-SRTM differencing. 
e An analysis of the number of ASTER individual scene 
DEMs that are stacked and averaged to derive the elevation 
value for every pixel in GDEM v2 shows that 
improvements to mean error and RMSE are minimal 
beyond about 10 to 15 scenes. 
e. GDEM v2 exhibits an apparent "true" negative elevation 
bias of about 1 meter, which was revealed through an 
analysis of mean error by land cover type. The overall 
mean error of —0.20 m (Figure 3 and Table 1) is certainly 
an improvement over the mean error of —3.69 for GDEM 
vl, but it somewhat masks the true performance of ASTER 
in measuring the elevation in open terrain conditions (non- 
vegetated, non-built-up). The overall mean error is 
dampened by the positive elevation biases contributed by 
forested and built-up land cover. While the true negative 
elevation bias of about 1 meter for GDEM v2 is a 
significant improvement over the true negative elevation 
bias of about 5 meters for GDEM vl, it is nonetheless a 
condition that users of GDEM v2 data should be aware of 
and factor into decisions regarding application of the 
product. 
5. REFERENCES 
Abrams, M., Bailey, B., Tsu, H., and Hato, M., 2010. The 
ASTER global DEM. Photogrammetric Engineering and 
Remote Sensing, 76(4), pp. 344-348. 
ASTER GDEM Validation Team, 2009. ASTER Global DEM 
Validation Summary Report. 28 p. 
http://www.ersdac.or.jp/GDEM/E/image/ASTERGDEM Valida 
tionSummaryReport Verl.pdf (5 Apr. 2012). 
ASTER GDEM Validation Team, 2011. ASTER Global Digital 
Elevation Model Version 2 — Summary of Validation Results. 
26 p. 
https://igskmncnwb001 .cr.usgs.gov/aste/ GDEM/Summary GD 
EM2 validation report final.pdf (5 Apr. 2012). 
Carabajal, C.C., and Harding, D.J., 2006. SRTM C-band and 
ICESat laser altimetry elevation comparisons as a function of 
tree cover and relief. Photogrammetric Engineering & Remote 
Sensing, 72(3), pp. 287-298. 
  
  
GDEM v2 GDEM v2 
Land cover Description mean error mean 
class vs. GPS difference 
benchmarks vs. NED 
  
Developed, e 
Open 
Space 
mostly lawn grasses, 
with some 
construction 
e <20% impervious 20:86: m 072m 
surfaces 
®  large-lot single-family 
housing units, parks, 
golf courses 
  
Developed, € 20-49% impervious 
  
  
  
Low 
Intensity surfaces A E 0.12 m 1.16 m 
® single-family housing 
units 
NEUES ® 50-79% impervious 
edium 
surfaces 
Intensity : ; i 0.79 m 1.48 m 
®  single-family housing 
units 
Developed, ® 80-100% impervious 
Hh : surfaces 
ntensi 
ty 9 apartment complexes, L77m 233m 
row houses, 
commercial/industrial 
  
  
  
  
Table 2. Increasing GDEM v2 mean error with increasing 
density of developed land cover. 
Farr, T.G., Rosen, P.A., Caro, E.. Crippen, R., Duren, R., 
Hensley, S., Kobrick, M., Paller, M., Rodriguez, E., Roth, L., 
Seal, D., Shaffer, S., Shimada, J., Umland, J., Werner, M., 
Oskin, M., Burbank, D., and Alsdorf, D., 2007. The Shuttle 
Radar  Topography Mission. Reviews of Geophysics, 
45(RG2004), doi:10.1029/2005RG000183. 
Fry, J.A., Xian, G., Jin, S., Dewitz, J.A., Homer, C.G., Yang, L., 
Barnes, C.A., Herold, N.D., and Wickham, J.D., 2011. 
Completion of the 2006 National Land Cover Database for the 
conterminous United States. Photogrammetric Engineering and 
Remote Sensing, 77(9), pp. 858-864. 
Gesch, D.B., 2007. The National Elevation Dataset. In: Maune, 
D. ed, Digital Elevation Model Technologies and 
Applications: The DEM Users Manual, 2" Edition. American 
Society for Photogrammetry and Remote Sensing, Bethesda, 
Maryland, pp. 99-118. 
Hofton, M., Dubayah, R., Blair, J.B., and Rabine, D., 2006. 
Validation of SRTM elevations over vegetated and non- 
vegetated terrain using medium footprint lidar. 
Photogrammetric Engineering & Remote Sensing, 72(3), pp. 
279-285. 
Hvidegaard, S.M., Soerensen, L.S., and Forsberg, R., 2012. 
ASTER GDEM validation using LiDAR data over coastal 
regions of Greenland. Remote Sensing Letters, 3(1), pp. 85-91. 
Maune, D.F., Maitra, J.B., and McKay, E.J., 2007. Accuracy 
standards & guidelines. In: Maune, D. (ed.), Digital Elevation 
Model Technologies and Applications: The DEM Users 
Manual, 2" Edition. American Society for Photogrammetry and 
Remote Sensing, Bethesda, Maryland, pp. 65-97. 
Miliaresis, G.Ch., and Paraschou, C.V.E., 2011. An evaluation 
of the accuracy of the ASTER GDEM and the role of stack