|Reference : Potentialities of using ASTER & SRTM DEM for road planning in Central African sustain...|
|Scientific congresses and symposiums : Poster|
|Life sciences : Phytobiology (plant sciences, forestry, mycology...)|
|Potentialities of using ASTER & SRTM DEM for road planning in Central African sustainable forest logging context. A case study in East Gabon|
|Philippart, Julien [Université de Liège - ULg > Forêts, Nature et Paysage > Gestion des ressources forestières et des milieux naturels >]|
|Handerek, Daphné [Université de Liège - ULg > Forêts, Nature et Paysage > Gestion des ressources forestières et des milieux naturels >]|
|Lejeune, Philippe [Université de Liège - ULg > Forêts, Nature et Paysage > Gestion des ressources forestières et des milieux naturels >]|
|Doucet, Jean-Louis [Université de Liège - ULg > Forêts, Nature et Paysage > Laboratoire de Foresterie des régions trop. et subtropicales >]|
|1° EARSeL SIG Forestry Workshop : Operational remote sensing in forest management|
|2-3 juin 2011|
|European Association of remote Sensing Laboratories|
|[en] Road planning ; slope ; ASTER and SRTM|
|[en] Slope is the main constraint for sustainable forest road planning in Central Africa. Remote sensing now provides free and downloadable Data Elevation Model (DEM) covering most of appeared lands. In this study, we evaluate potentialities and limitations of Shuttle Radar Topography Mission (SRTM) DEMs, derived from radar interferometry and Advanced Spaceborn Thermal Emission and Reflection Radiometer (ASTER) DEMs, themselves derived from digital photogrammetry for slope evaluation. Assessment is presented for hilly (Milole) and flat (Okondja) areas. Three elevation maps were derived from initial ASTER 30-m and SRTM 90-m DEMs : a SRTM 30-m resampled from SRTM 90-m and two ASTER 30-m where absurd values (artifacts) were corrected with SRTM 90-m and resampled SRTM 30-m respectively. We qualitatively and quantitatively assess the accuracy of all elevation maps compared to 992 (698) slope measures on field along transects of 10.5 (7.5) km in Milolé (Okondja).
We estimated root mean square error (RMSE) for slope estimation at 7.8 (10.7), 8.1 (10.1), 11.7 (11.2), 10.1 (11.2) and, 9.3° (11.0°) for SRTM 90-m, SRTM 30-m, ASTER 30-m, ASTER 30-m CORR 90 and ASTER 30-m CORR 30 respectively in Milolé (Okondja).
We also use a classification error matrix to assess Global Accuracy (GA) and User’s Accuracy (UA) of elevation maps by classifying ground slopes in two categories: lower or equal and higher than maximum slope limitation of 12% (30%) for primary (secondary) roads.
All DEMs show a greater GA in hilly area (Milolé) than in flat area (Okondja) and SRTM derived DEMs show a higher UA for secondary roads constraint.
UA for the lower or equal category varies between 55.5 and 75.2% (63.9 and 91.7%) for primary (secondary) roads.
The use of corrected aster DEMs increases initial ASTER UA from 0.1 to 18.8% depending on category and slope limitation.
Despite a relatively high RMSE for slope grade, all of the DEMs tested were found to be consistent for consideration of maximum slope constraint aiming at sustainable road planning for forest logging in Central Africa.
|Researchers ; Professionals ; Students ; General public ; Others|
|File(s) associated to this reference|
All documents in ORBi are protected by a user license.