There is a newer version of this record available.

Dataset Open Access

Continental Europe Digital Terrain Model at 30 m resolution based on GEDI and background layers

Hengl, Tomislav; Leal Parente, Leandro; Krizan, Josip

DataCite XML Export

<?xml version='1.0' encoding='utf-8'?>
<resource xmlns:xsi="" xmlns="" xsi:schemaLocation="">
  <identifier identifierType="DOI">10.5281/zenodo.4056635</identifier>
      <creatorName>Hengl, Tomislav</creatorName>
      <nameIdentifier nameIdentifierScheme="ORCID" schemeURI="">0000-0002-9921-5129</nameIdentifier>
      <affiliation>OpenGeoHub foundation</affiliation>
      <creatorName>Leal Parente, Leandro</creatorName>
      <familyName>Leal Parente</familyName>
      <nameIdentifier nameIdentifierScheme="ORCID" schemeURI="">0000-0003-1589-0467</nameIdentifier>
      <affiliation>OpenGeoHub foundation</affiliation>
      <creatorName>Krizan, Josip</creatorName>
      <nameIdentifier nameIdentifierScheme="ORCID" schemeURI="">0000-0002-4557-3537</nameIdentifier>
    <title>Continental Europe Digital Terrain Model at 30 m resolution based on GEDI and background layers</title>
    <subject>digital terrain model</subject>
    <subject>ensemble machine learning</subject>
    <subject>elevation data</subject>
    <date dateType="Issued">2020-09-28</date>
  <resourceType resourceTypeGeneral="Dataset"/>
    <alternateIdentifier alternateIdentifierType="url"></alternateIdentifier>
    <relatedIdentifier relatedIdentifierType="DOI" relationType="IsVersionOf">10.5281/zenodo.4056634</relatedIdentifier>
    <rights rightsURI="">Creative Commons Attribution 4.0 International</rights>
    <rights rightsURI="info:eu-repo/semantics/openAccess">Open Access</rights>
    <description descriptionType="Abstract">&lt;p&gt;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&amp;nbsp;GEDI&amp;nbsp;level 2B points (column &amp;quot;elev.lowestmode&amp;quot;): about 7 million GEDI points were overlaid vs NASADEM, AW3D, EU DEM, canopy height, tree cover and surface water cover maps, then a 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:&lt;/p&gt;

	&lt;li&gt;&amp;quot;lcv_bare.earth_glcf.landsat&amp;quot;: UMD GLAD&amp;nbsp;&lt;a href=""&gt;bare earth estimate&lt;/a&gt; for year 2010 based on Landsat time series,&lt;/li&gt;
	&lt;li&gt;&amp;quot;dtm_elev.dsm_alos.aw3d&amp;quot;: Digital Surface Model based on &lt;a href=""&gt;ALOS AW3D v8104&lt;/a&gt;,&amp;nbsp;&lt;/li&gt;
	&lt;li&gt;&amp;quot;dtm_canopy.height_glad.umd&amp;quot;: UMD GLAD &lt;a href=""&gt;canopy height for 2019&lt;/a&gt; based on GEDI data,&lt;/li&gt;
	&lt;li&gt;&amp;quot;dtm_elev.dsm_eudem.eea&amp;quot;: Copernicus&amp;nbsp;&lt;a href=""&gt;EU DEM&lt;/a&gt; based on the SRTM and ASTER DEMs,&lt;/li&gt;
	&lt;li&gt;&amp;quot;hyd_surface.water_jrc.gswe&amp;quot;: &lt;a href=""&gt;JRC Global Surface Water Explorer&lt;/a&gt;&amp;nbsp;surface water probability based on the Landsat time-series,&lt;/li&gt;
	&lt;li&gt;&amp;quot;dtm_elev.dsm_nasadem.hgt&amp;quot;: Digital Surface Model based on the &lt;a href=""&gt;USGS NASADEM&lt;/a&gt;,&lt;/li&gt;
	&lt;li&gt;&amp;quot;lcv_landcover.12_pflugmacher2019&amp;quot;: land cover map of Europe at 30 based on &lt;a href=""&gt;Pflugmacher et al. (2019)&lt;/a&gt;,&lt;/li&gt;
	&lt;li&gt;&amp;quot;lcv_tree.cover_umd.landsat_2000&amp;quot;: forest tree cover for year 2000 based on the &lt;a href=""&gt;Global Forest Change data&lt;/a&gt;,&lt;/li&gt;
	&lt;li&gt;&amp;quot;lcv_tree.cover_umd.landsat_2010&amp;quot;:&amp;nbsp;forest tree cover for year 2010 based on the &lt;a href=""&gt;Global Forest Change data&lt;/a&gt;,&lt;/li&gt;

&lt;p&gt;Detailed processing steps can be found &lt;a href=""&gt;&lt;strong&gt;here&lt;/strong&gt;&lt;/a&gt;. Summary results of the model training (&lt;a href=""&gt;mlr::makeStackedLearner&lt;/a&gt;) report:&lt;/p&gt;

stats::lm(formula = f, data = d)

    Min      1Q  Median      3Q     Max 
-65.580  -2.630   0.648   3.120 181.769 

                Estimate Std. Error t value Pr(&amp;gt;|t|)    
(Intercept)   -4.1448129  0.4663283  -8.888  &amp;lt; 2e-16 ***
regr.ranger    0.2667469  0.0009676 275.677  &amp;lt; 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  &amp;lt; 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: &amp;lt; 2.2e-16&lt;/code&gt;&lt;/pre&gt;

&lt;p&gt;The output predicted terrain model includes the following two layers:&lt;/p&gt;

	&lt;li&gt;&amp;quot;dtm_elev.lowestmode_gedi.eml_m&amp;quot;: mean estimate of the terrain elevation,&lt;/li&gt;
	&lt;li&gt;&amp;quot;dtm_elev.lowestmode_gedi.eml_md&amp;quot;: standard deviation of the independently fitted stacked predictors quantifying the prediction uncertainty,&lt;/li&gt;

&lt;p&gt;The predicted elevations are based on the GEDI data hence the &lt;strong&gt;reference water surface (WGS84 ellipsoid) is about &lt;a href=""&gt;43 m higher&lt;/a&gt; than the sea water surface for a specific EU country&lt;/strong&gt;. All GeoTIFFs were prepared using Integer format (elevations rounded to 1 m) and have been converted to &lt;a href=""&gt;Cloud Optimized GeoTIFFs using GDAL&lt;/a&gt;.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;Disclaimer&lt;/strong&gt;: The output DTM shows forest tops 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 &lt;a href=""&gt;&lt;strong&gt;here&lt;/strong&gt;&lt;/a&gt;. To register for updates please subscribe to: &lt;a href=""&gt;;/a&gt;.&lt;/p&gt;</description>
    <description descriptionType="Other">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 (</description>
    <description descriptionType="Other">{"references": ["Bischl, B., Lang, M., Kotthoff, L., Schiffner, J., Richter, J., Studerus, E., ... &amp; Jones, Z. M. (2016). mlr: Machine Learning in R. The Journal of Machine Learning Research, 17(1), 5938-5942.", "Hansen, M. C., Potapov, P. V., Moore, R., Hancher, M., Turubanova, S. A., Tyukavina, A., ... &amp; Kommareddy, A. (2013). High-resolution global maps of 21st-century forest cover change. science, 342(6160), 850-853.", "Pekel, J. F., Cottam, A., Gorelick, N., &amp; Belward, A. S. (2016). High-resolution mapping of global surface water and its long-term changes. Nature, 540(7633), 418-422.", "Pflugmacher, D., Rabe, A., Peters, M., &amp; Hostert, P. (2019). Mapping pan-European land cover using Landsat spectral-temporal metrics and the European LUCAS survey. Remote sensing of environment, 221, 583-595.", "Potapov, P., X. Li, A. Hernandez-Serna, A. Tyukavina, M.C. Hansen, A. Kommareddy, A. Pickens, S. Turubanova, H. Tang, C.E. Silva, J. Armston, R. Dubayah, J. B. Blair, M. Hofton (2020) Mapping and monitoring global forest canopy height through integration of GEDI and Landsat data. In review", "Takaku, J., &amp; Tadono, T. (2017). Quality updates of 'AW3D'global DSM generated from ALOS PRISM. In 2017 IEEE International Geoscience and Remote Sensing Symposium (Igarss) (pp. 5666-5669). IEEE."]}</description>
All versions This version
Views 1,30522
Downloads 4573
Data volume 2.1 TB5.4 GB
Unique views 1,11222
Unique downloads 2213


Cite as