• Aucun résultat trouvé

DEM smoothing to avoid the staircase effect of coarse DEM

N/A
N/A
Protected

Academic year: 2022

Partager "DEM smoothing to avoid the staircase effect of coarse DEM"

Copied!
2
0
0

Texte intégral

(1)

HAL Id: hal-01209158

https://hal.archives-ouvertes.fr/hal-01209158

Submitted on 6 Jun 2020

HAL is a multi-disciplinary open access archive for the deposit and dissemination of sci- entific research documents, whether they are pub- lished or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.

DEM smoothing to avoid the staircase effect of coarse DEM

Sébastien Lehmann, Hervé Squividant, Nicolas Saby, Jean-Baptiste Paroissien, Rossano Ciampalini, Manuel Martin

To cite this version:

Sébastien Lehmann, Hervé Squividant, Nicolas Saby, Jean-Baptiste Paroissien, Rossano Ciampalini, et al.. DEM smoothing to avoid the staircase effect of coarse DEM. Global Soil Map, Oct 2013, Orléans, France. pp.1. �hal-01209158�

(2)

DEM smoothing to avoid the staircase effect of coarse DEM

Lehmann S.

1

, Squividant H.

2

, Saby N.P.A

1

, Paroissien J-B

1

, Daroussin J.

3

, Ciampalini R.

1

, Martin, M.P.

1

1INRA, US 1106, InfoSol Unit, CS 40001 Ardon, F-45075 Orleans 2, France

2INRA, UMR1069, Sol Agro et hydrosystème Spatialisation, F-35000 Rennes, France

3INRA, UR0272, UR Science du Sol, CS 40001 Ardon, F-45075 Orleans 2, France 15km

INTRODUCTION

MATERIAL & METHOD

CONCLUSION - PERSPECTIVES

RESULTS

Slope

(degree)

0 - 0,001

0,001 - 0,002 0,002 - 0,05 0,051 - 0,06 0,061 - 0,1 0,101 - 0,2 0,201 - 0,5 0,501 - 1 1,001 - 3 3,001 - 10

Elevation original DEM IDW SPLINE TIN MNT_F measured 0,904 0,919 0,917 0,911 0,935

The original DEM shows a majority of its surface (75%) covered with slopes equal to 0. The proportion of null slopes decreases with the following interpolations : IDW (66%), Spline (34%), TIN (23%), MNT_F (12%).

The MNT_F interpolation gives a majority of slopes (25%) between 0,2 and 0,5 degree.

Original DEM IDW SPLINE TIN MNT_F

Area :

225 km²

Parent Material : Recent alluvial deposit from the Rhine (Holocene) Morphology :

Rhine valley with a very low slope

oriented to the NE.

Land-use :

Intensive agri & forest Climate :

Semi-continental

Mean annual T : 10° C Total annual

precipitation : 600mm

Measured

MNT_F

Pearson product-moment correlation coefficient for all interpolations compared with the measured elevations

Original DEM Elevation (m)

high : 191 low : 154

TIN Interpolation Elevation (m)

high : 191 low : 154 Isolines of Elevations (m)

In the case of delineation of wetlands, the Terrain Wetness Index (Beven and Kirkby, 1979) is particularly relevant but also sensitive to flat areas. The altitudinal resolution also has a strong influence on variable because it affects directly slope calculation. A metrical resolution in elevation for a 25 m DEM (horizontal resolution), as provided by the IGN (French National Geographic Institute), leads inevitably to a staircase effect with leaps over large surfaces which wrongly appear as flat and straight.

The study shown here compares different methods of DEM smoothing which can be used to mitigate the staircase effect and produce a more realistic model with a decimetric altitudinal resolution.

Study area located

in North-East France, In Alsace region

We show that using a triangular irregular network interpolation, plus hard lines (streams and orographic lines) as calculated by the Terrain function in ArcGis, significantly improves the suitability of resulting DEM to calculate the TWI. To finish the demonstration, this results should be compared with observations of wet soils in the same region and with a more accurate DEM (e.g. LiDAR DEM).

References

Beven, K.J., Kirkby, M.J., 1979. A physically based, variable contributing area model of basin hydrology. Hydrological Sciences Bulletin 24 (1), p. 43–69.

Franke, R. 1982. Smooth Interpolation of Scattered Data by Local Thin Plate Splines. Computer and Mathematics with Appllications. Vol. 8. No. 4. p. 273–281.

Herbert Edelsbrunner, 1996. « Incremental Topological Flipping Works for Regular Triangulations », Algorithmica, vol. 15, p. 223–241.

Software

ESRI 2011. ArcGIS Desktop: Release 10. Redlands, CA: Environmental Systems Research Institute

Measured vs interpolated elevations

The original DEM is converted into isolines of Elevations or single points situated on those lines. The elevations data from the original DEM are then interpolated using several methods implemented in ArcGis (ESRI, 2011) :

Inverse Distance Weighted interpolation (IDW), Spline function with tension (Franck, 1982) (SPLINE), Triangular Irregular Networks (Edelsbrunner, 1996) (TIN), Terrain function integrating topographic breaklines (MNT_F). The interpolations are validated by a dataset of 60 geodetic points measured by the IGN with centimetric precision.

TWI value

-9 - -4 -3 - 1 2 3 - 4 5 6 7 - 8 9 - 11 12 - 16 17 - 21

IDW SPLINE TIN MNT_F

No data (Elevation) Original DEM

Contact : [email protected]

The Terrain Wetness Index (TWI) is directly correlated with the slope gradient. It appears that the TIN and MNT_F interpolations improve consequently the quality of this index. The MNT_F which takes into account the orographic and streams lines with their real elevations (given by the IGN BD_TOPO) is obvioulsly the more realistic.

The accuracy of the index calculation is validated with the Pearson coefficient comparing interpolated elevations and the elevations measured on the geodetic points.

y = 1,0012x

154 156 158 160 162 164 166 168 170 172

155 160 165 170 175

# Measured altitude

#

#

#

#

#

# #

#

# # #

#

#

# #

# #

#

#

#

#

#

#

#

# #

#

#

#

#

#

#

#

#

# #

# #

#

# #

# #

#

#

#

#

#

#

#

# #

#

#

#

# #

#

#

#

#

#

#

Références

Documents relatifs

L’objectif principal de notre étude était d’étudier la CRP à l’admission comme facteur pronostique chez les patients admis dans les services de médecine interne via les

Enfin, la description de chaque article du corpus de références de la base PASCAL (65066 références) par la classification de l’ESI revient à construire la table de

Pour les régions des PECO, alors que seule Prague était dans le cadran HB en 2000, on note qu’en 2006 on trouve les régions po- lonaises de Varsovie et Lodz (respectivement 1 ère

Les fondateurs de la maison Mimir l’ont donc investie et ont très vite communiqué avec la mairie de Strasbourg : « Nous on s’est installés, on a écrit une lettre à la mairie

L’analyse du réseau global des relations entre tous les stagiaires PE2 de l’IUFM montre que les groupes de référence sont des lieux d’échange et de relations parfois

Nicolas Saby, Agnès Bourgeois, Denis Allard, Hakima Boukir, Manuel Martin, Marion Bardy, Dominique Arrouays.. To cite

Fanny Collard, Bas Kempen, Gerard Bm Heuvelink, Nicolas Saby, Anne C Richer-De-Forges, Sébastien Lehmann, Pierre Nehlig, Dominique Arrouays.. To cite

DURÉE DE LA PÉRIODE DE REPRODUCTION CHEZ LES FEMELLES DE LONISCOÏDE POR- CELLIO DILATATUS BRANDT SUIVANT LES CONDITIONS D’ÉLEVAGE : TEMPÉRATURE, PHOTOPÉRIODE ET GROUPEMENT... DURÉE