4057883
doi
10.5281/zenodo.4057883
oai:zenodo.org:4057883
Leal Parente, Leandro
OpenGeoHub foundation
Krizan, Josip
MultiOne
Bonannella, Carmelo
OpenGeoHub foundation
Continental Europe Digital Terrain Model at 30 m resolution based on GEDI and background layers
Hengl, Tomislav
OpenGeoHub foundation
doi:10.5281/zenodo.4058447
info:eu-repo/semantics/openAccess
Creative Commons Attribution 4.0 International
https://creativecommons.org/licenses/by/4.0/legalcode
digital terrain model
Europe
geomorphometry
ensemble machine learning
GEDI
elevation data
<p>Digital Terrain Model for Continental Europe based on the three publicly available Digital Surface Models and predicted using an Ensemble Machine Learning (EML). EML was trainined using GEDI level 2B points (column "elev.lowestmode"): about 7 million GEDI points were overlaid vs NASADEM, AW3D, EU DEM, canopy height, tree cover and surface water cover maps, then an ensemble prediction model was fitted using random forest, GLM with Lasso, Cubist and GLMnet, and used to predict most probable terrain height (bare earth). Input layers used to train the EML include:</p>
<ul>
<li>"lcv_bare.earth_glcf.landsat": UMD GLAD <a href="http://glad.umd.edu/dataset/global-2010-bare-ground-30-m">bare earth estimate</a> for year 2010 based on Landsat time series,</li>
<li>"dtm_elev.dsm_alos.aw3d": Digital Surface Model based on <a href="https://www.eorc.jaxa.jp/ALOS/en/aw3d/index_e.htm">ALOS AW3D</a>, </li>
<li>"dtm_canopy.height_glad.umd": UMD GLAD <a href="http://glad.umd.edu/dataset/gedi/">canopy height for 2019</a> based on GEDI data,</li>
<li>"dtm_elev.dsm_eudem.eea": Copernicus <a href="https://www.eea.europa.eu/data-and-maps/data/copernicus-land-monitoring-service-eu-dem">EU DEM</a> based on the SRTM and ASTER DEMs,</li>
<li>"hyd_surface.water_jrc.gswe": <a href="https://global-surface-water.appspot.com/">JRC Global Surface Water Explorer</a> surface water probability based on the Landsat time-series,</li>
<li>"dtm_elev.dsm_nasadem.hgt": Digital Surface Model based on the <a href="https://lpdaac.usgs.gov/products/nasadem_hgtv001/">USGS NASADEM</a>,</li>
<li>"lcv_landcover.12_pflugmacher2019": land cover map of Europe at 30 based on <a href="https://doi.pangaea.de/10.1594/PANGAEA.896282">Pflugmacher et al. (2019)</a>,</li>
<li>"lcv_tree.cover_umd.landsat_2000": forest tree cover for year 2000 based on the <a href="https://earthenginepartners.appspot.com/science-2013-global-forest">Global Forest Change data</a>,</li>
<li>"lcv_tree.cover_umd.landsat_2010": forest tree cover for year 2010 based on the <a href="https://earthenginepartners.appspot.com/science-2013-global-forest">Global Forest Change data</a>,</li>
</ul>
<p>Detailed processing steps can be found <a href="https://gitlab.com/geoharmonizer_inea/spatial-layers"><strong>here</strong></a>. Read more about the processing steps <a href="https://opendatascience.eu/building-continental-europe-digital-terrain-model-30-m-resolution-using-machine-learning"><strong>here</strong></a>. Training data set can be obtained in the file "gedi_elev.lowestmode_2019_eumap.RDS". Summary results of the model training (<a href="https://mlr.mlr-org.com/reference/makeStackedLearner.html">mlr::makeStackedLearner</a>) report:</p>
<pre><code>Call:
stats::lm(formula = f, data = d)
Residuals:
Min 1Q Median 3Q Max
-65.580 -2.630 0.648 3.120 181.769
Coefficients:
Estimate Std. Error t value Pr(>|t|)
(Intercept) -4.1448129 0.4663283 -8.888 < 2e-16 ***
regr.ranger 0.2667469 0.0009676 275.677 < 2e-16 ***
regr.glmnet -4.7183974 0.6038334 -7.814 5.54e-15 ***
regr.cvglmnet 4.6966219 0.6042481 7.773 7.69e-15 ***
regr.cubist 0.7643997 0.0012860 594.378 < 2e-16 ***
---
Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
Residual standard error: 6.729 on 6757726 degrees of freedom
Multiple R-squared: 0.9996, Adjusted R-squared: 0.9996
F-statistic: 4.644e+09 on 4 and 6757726 DF, p-value: < 2.2e-16</code></pre>
<p>Which indicates that the elevation errors are in average (2/3rd of pixels) between +2-3 m. The output predicted terrain model includes the following two layers:</p>
<ul>
<li>"dtm_elev.lowestmode_gedi.eml_m": mean estimate of the terrain elevation,</li>
<li>"dtm_elev.lowestmode_gedi.eml_md": standard deviation of the independently fitted stacked predictors quantifying the prediction uncertainty,</li>
</ul>
<p>The predicted elevations are based on the GEDI data hence the <strong>reference water surface (WGS84 ellipsoid) is about <a href="https://kartoweb.itc.nl/geometrics/Reference%20surfaces/refsurf.html">43 m higher</a> than the sea water surface for a specific EU country</strong>. All GeoTIFFs were prepared using Integer format (elevations rounded to 1 m) and have been converted to <a href="https://gdal.org/drivers/raster/cog.html">Cloud Optimized GeoTIFFs using GDAL</a>.</p>
<p><strong>Disclaimer</strong>: The output DTM still shows forest canopy (overestimation of the terrain elevation) and has not been hydrologically corrected for spurious sinks and similar. This data set is continuously updated. To report a bug or suggest an improvement, please visit <a href="https://gitlab.com/geoharmonizer_inea/spatial-layers/-/issues"><strong>here</strong></a>. To register for updates please subscribe to: <a href="https://twitter.com/HarmonizerGeo">https://twitter.com/HarmonizerGeo</a>.</p>
This work has received funding from the European Union's the Innovation and Networks Executive Agency (INEA) under Grant Agreement Connecting Europe Facility (CEF) Telecom project 2018-EU-IA-0095 (https://ec.europa.eu/inea/en/connecting-europe-facility/cef-telecom/2018-eu-ia-0095).
Zenodo
2020-09-28
info:eu-repo/semantics/other
4056634
v0.1
1619622336.375962
7971909541
md5:f2f08cf61704c5969ff55d24cfb220e8
https://zenodo.org/records/4057883/files/dtm_elev.lowestmode_gedi.eml_m_30m_0..0cm_2000..2018_eumap_epsg3035_v0.1.tif
1449150
md5:0f960837d2385943460e1721a82ba131
https://zenodo.org/records/4057883/files/001_preview_EU_DTM.png
8029447000
md5:1c71a8782a4da623e6883d041b89f0b3
https://zenodo.org/records/4057883/files/dtm_elev.dsm_eudem.eea_m_30m_0..0cm_2002..2014_eumap_epsg3035_v1.0.tif
3696999260
md5:b7136655458433c798b064d22d089b69
https://zenodo.org/records/4057883/files/dtm_canopy.height_glad.umd_m_30m_0..0cm_2019_eumap_epsg3035_v0.1.tif
1549831591
md5:7e3e3e2e398fa5fa9594a930d6e1095e
https://zenodo.org/records/4057883/files/gedi_elev.lowestmode_2019_eumap.RDS
998194349
md5:af37fef0c0e8ae693888c5bc737f90a5
https://zenodo.org/records/4057883/files/hyd_surface.water_jrc.gswe_p_30m_0..0cm_1984..2019_eumap_epsg3035_v0.1.tif
623938706
md5:a1fede759ea9ec9c76169a31b15131ac
https://zenodo.org/records/4057883/files/dtm_elev.lowestmode_gedi.eml_m_100m_0..0cm_2000..2018_eumap_epsg3035_v0.1.tif
13936409928
md5:32692709cfbff1942a4020fa26b57719
https://zenodo.org/records/4057883/files/dtm_elev.dsm_alos.aw3d_m_30m_0..0cm_2016..2018_eumap_epsg3035_v1804.tif
2718295186
md5:9451519c2f6bc98804a0bddc9956298b
https://zenodo.org/records/4057883/files/dtm_elev.lowestmode_gedi.eml_md_30m_0..0cm_2000..2018_eumap_epsg3035_v0.1.tif
4323667068
md5:31ac1244c16de3a59c3e7f8b25769273
https://zenodo.org/records/4057883/files/lcv_tree.cover_umd.landsat_m_30m_0..0cm_2010_eumap_epsg3035_v1.7.tif
1332398033
md5:e292d4800e8c8c4df40dea4b7f756478
https://zenodo.org/records/4057883/files/lcv_landcover.12_pflugmacher2019_c_1m_s0..0m_2014..2016_eumap_epsg3035_v0.1.tif
3495477581
md5:a4554a120b89bf3230b8e1373ca0c7b8
https://zenodo.org/records/4057883/files/lcv_bare.earth_glcf.landsat_m_30m_0..0cm_2010_eumap_epsg3035_v0.1.tif
public
10.5281/zenodo.4058447
Is supplemented by
doi
10.5281/zenodo.4056634
isVersionOf
doi