Hi Carsten, It is almost certain that you are right.
The chapter "C. Residual Anomalies and Artifacts GDEM" detailed "Artifacts Related to Stack Number Irregular Boundaries." - Decripiton of the "pits" is similar to more or less circular holes I see. The text says they "Regularity and Often Occur with high frequency in Virtually all ASTER GDEM tiles" and "from just a FEW meters to 100 meters or more." Ok - The large positive anomalies similar to "mole run" artifacts (p.23 of the text). - Undoubtedly the linear artifact are "step anomaly" (p. 19 text). Thank you for your comments, I'll put them on the blog to let things clear. Bests PS: as you assume, unfortunately, no amazon mystery reveled... Sylvain 2011/8/26 Carsten Troelsgaard <troelsgaard...@live.dk> > > I looked around for information about the artefacts in the aster data. The > Quality Assessment file that accompanies the data mentions: > Quote > The automated cloud masking and statistical approach used to select data for > stacking > are not totally effective in avoiding anomalous elevations values, and > anomalies may > remain in the GDEM where the stack number is three or less, particularly. > Where > available, existing DEMs were used to replace anomalous GDEM values, including > adjusting for offsets between the ASTER GDEM and the reference DEM data. > UnQuote > following the link: > http://www.ersdac.or.jp/GDEM/E/image/ASTERGDEM_ValidationSummaryReport_Ver1.pdf > At page ~20 the chapter "Artifacts Related to Irregular Stack Number > Boundaries." has additional information and visual displays equal to the ones > you link to at //yepka.. > It would be interesting if the artefacts reveal some secret about the > landscape, but so far I've treated as a measurement/processing-error. > > I assume that 'stack-number' originates from overlapping measurement-swats > > Carsten > ________________________________ > Sylv wrote: > > I used a re-sampling to 30m of SRTM-DEM (made by Brazilian INPE) and observe > a visual superposition between GRASS-r.watershed river network output and > 2001 Landsat7 deforested areas (the nearest date after SRTM) (see > http://yepca.org/wp3/?p=349) > As you suggest, I'll verify if CGIAR SRTM-DEM is better in such flat area > with forest/cultivation patchwork. > In first analyze, ASTER GDEM gave better river network, even with the theses > 20-30m holes or larges areas artifacts!... but more job is necessary to > verify and understand why! > > Sylv > > > > _______________________________________________ > Discuss mailing list > Discuss@lists.osgeo.org > http://lists.osgeo.org/mailman/listinfo/discuss > _______________________________________________ Discuss mailing list Discuss@lists.osgeo.org http://lists.osgeo.org/mailman/listinfo/discuss