There is a newer version of the record available.

Published February 9, 2025 | Version 1.0
Model Open

Global Ensemble Digital Terrain Model (GEDTM30): a Tile Example of Local Enhanced Modeling via Transfer Learning

Authors/Creators

  • 1. OpenGeoHub

Description

This repository showcases an example to run a local enhanced modeling via transfer learning in Python. The script describes a pre-trained random forest model (global_model_legendtm.lz4) that is successively trained by the additional local samples (local_sample.csv). With the data from (covariates list), global and local prediction and stanard deviation has been derived in files (gendtm_pred_std.tiff, gendtm_prediction.tiff, legendtm_prediction.tiff, legendtm_pred_std.tiff). The standard deviation is scaled 100 and saved in INT16 and prediciton is scaled 10 and saved in INT32. The follow-up section explains the files and their references.

 

File description:

global_model_legendtm.lz4: random forest model trained by global stratified terrain samples of GEDI02 and ICESat-2 ATL08 v6.

local_sample.csv: the local samples of GEDI02 and ICESat-2 ATL08 v6. dtm_y: the terrain height from global Lidar

covariates list:

  • GLO30_30m_v01272022.tiff: Copenricus GLO30 DEM (doi.org/10.5270/ESA-c5d3d65
  • dtm_elevation_aw3d30.dsm_m_30m_v01202022.tiff: ALOS AW3D30 DEM (Takaku et al. 2014, Tadono et al.)
  • ghs_built_h_anbh_e2018_globe_r2022a_54009_100_v1_0.tiff: GHS building height ~100m resolution, resampled by cubicspline (10.2905/85005901-3A49-48DD-9D19-6261354F56FE)
  • tree.cover_glad.umd_m_30m_s_20100101_20151231_go_epsg.4326_v1.0.tiff: tree cover precentile (Potapov et al. 2021)
  • 3d_globfp.tiff: global three-dimensional building footprint dataset, vector dataset, rasterized to ~30m resolution (Che et al. 2024)
  • canopy.height_glad.umd_m_30m_s_2017_2018_go_epsg.4326_v01242022.tiff: canopy height from GLAD (Potapov et al. 2021)
  • edge.canopy.height_glad.umd_m_30m_s_2017_2018_go_epsg.4326_v01242022.tiff: edge canopy height by Sobel filter, derived from GLAD canopy height 
  • canopy.height_eth_m_30m_s_20000101_20230101_go_epsg.4326_v01242022.tiff: canopy height model from ETH (Lang et al. 2023)
  • edge.canopy.height_eth_m_30m_s_20000101_20230101_go_epsg.4326_v01242022.tiff:edge canopy height by Sobel filter, derived from ETH canopy height
  • ETOPO_slope_2022.tiff: earth topography 2022 (etopo 2022) global dem dataset ~500m resolution, resampled by cubicspline (MacFerrin et al. 2024)
  • lc_glad.glcluc.change_c_30m_s_20000101_20201231_go_epsg.4326_v20230901.tiff: land use land cover change (Potapov et al. 2020)
  • Landsat indices (Potapov 2020):

ndvi_swa.ard2_p025_30m_s_20060101_20101231_go_epsg.4326_v1.in.tiff
ndvi_swa.ard2_p50_30m_s_20060101_20101231_go_epsg.4326_v1.in.tiff
ndvi_swa.ard2_p975_30m_s_20060101_20101231_go_epsg.4326_v1.in.tiff
ndvi_swa.ard2_p025_30m_s_20110101_20151231_go_epsg.4326_v1.in.tiff
ndvi_swa.ard2_p50_30m_s_20110101_20151231_go_epsg.4326_v1.in.tiff
ndvi_swa.ard2_p975_30m_s_20110101_20151231_go_epsg.4326_v1.in.tiff
nir_swa.ard2_p025_30m_s_20060101_20101231_go_epsg.4326_v1.in.tiff
nir_swa.ard2_p50_30m_s_20060101_20101231_go_epsg.4326_v1.in.tiff
nir_swa.ard2_p975_30m_s_20060101_20101231_go_epsg.4326_v1.in.tifff
nir_swa.ard2_p025_30m_s_20110101_20151231_go_epsg.4326_v1.in.tiff
nir_swa.ard2_p50_30m_s_20110101_20151231_go_epsg.4326_v1.in.tiff
nir_swa.ard2_p975_30m_s_20110101_20151231_go_epsg.4326_v1.in.tif
ndwi_swa.ard2_p025_30m_s_20060101_20101231_go_epsg.4326_v1.in.tiff
ndwi_swa.ard2_p50_30m_s_20060101_20101231_go_epsg.4326_v1.in.tiff
ndwi_swa.ard2_p975_30m_s_20060101_20101231_go_epsg.4326_v1.in.tiff
ndwi_swa.ard2_p025_30m_s_20110101_20151231_go_epsg.4326_v1.in.tiff
ndwi_swa.ard2_p50_30m_s_20110101_20151231_go_epsg.4326_v1.in.tiff
ndwi_swa.ard2_p975_30m_s_20110101_20151231_go_epsg.4326_v1.in.tiff

nir,ndvi,ndwi: Near infra-red, Normalized Difference Vegetation Index, Normalized Difference Wetness Index

p025, p050,p975: precentile 2.5, 50, 97.5.

predictions:

  • gendtm_predictiontiff: global model prediction
  • gendtm_pred_std.tiff: standard deivation of global model prediction 
  • legendtm_pred_std.tiff: local enhanced global model prediction
  • legendtm_pred_std.tiff: standard deivation of local enhanced global model prediction

Files

3d_globfp.tiff

Files (11.7 GB)

Name Size
md5:fbd60f2f35b2ca86229ec8b6f298ce05
1.9 kB Preview Download
md5:27f0499480ab6dd2e424052b1ea0a989
125.3 MB Preview Download
md5:a876869dd08211296ceecf8d8ad8bc63
93.6 MB Preview Download
md5:6576a2ef6713c24e3388fe9f4512776b
348.4 MB Preview Download
md5:1b86036b39ce454a7f1846286da97a12
457.1 MB Preview Download
md5:31694545582fe71523fcbc0da033bab4
373.9 MB Preview Download
md5:937a206dfe529a5907e22e2f55768a83
421.5 MB Preview Download
md5:7ce22c4f663079419848daadcf736577
400.5 MB Preview Download
md5:b091aababd78227ca917fc5fb62a9b91
678.1 MB Preview Download
md5:54a1aefb75f6ee50b3cd6071544f5de6
811.5 MB Preview Download
md5:335520601e0b1759faea9c5aea331fef
338.2 MB Preview Download
md5:a484a81742908a142f3d79d539f799ba
107.4 MB Preview Download
md5:59e3cf08823d36dbb6b7b9a3812e5a74
567.7 MB Preview Download
md5:e224cad24e189fe1fdc62467851f83c3
679.1 MB Preview Download
md5:16dee413855bf3765bce446c9d544a55
297.9 kB Preview Download
md5:cdca201332d32b1c9db7091d203e71d6
54.6 MB Preview Download
md5:6a6133aeb5597682c9d270c7509b922e
2.3 GB Download
md5:235e39af3379c2e3370474258488e7db
217.5 MB Preview Download
md5:7479d27e4833fd3f4497c54a922b6dda
226.9 MB Preview Download
md5:c1eb4a4052ca3e5ef61e353339b0ec6c
196.8 MB Preview Download
md5:ca49fa855e2d2a559bf9b60bc866d8b5
203.0 MB Preview Download
md5:419229d2b07eb71bee884891d0eb27f4
207.0 MB Preview Download
md5:c935cbebcdc71ad762ba22ff3657c713
210.7 MB Preview Download
md5:49a866a10e9b56a9871d5881ab78f77f
208.2 MB Preview Download
md5:bf9d14d7ff57bf43dec0dfd96f10d9f7
218.1 MB Preview Download
md5:821b0632149bbeede4d1361c2acbb754
192.7 MB Preview Download
md5:d2efe41b48eefa52ab9a62816d0ef111
198.1 MB Preview Download
md5:a2ead790ed24e89a9800b138d215f626
209.8 MB Preview Download
md5:e475d980cddacea9cbee7aa7670957cc
215.5 MB Preview Download
md5:a472866f4d42bd4f181518602e6f6d76
200.6 MB Preview Download
md5:ffd6007e4b9d9a4a5546c1d07b9d3859
209.4 MB Preview Download
md5:8c351a9b77ae1a70d9021e0419afb4c5
187.0 MB Preview Download
md5:1004e643ff11be7b9b44c4c871626864
194.0 MB Preview Download
md5:734f5c28a394ec59c7d0437f56e5d827
214.4 MB Preview Download
md5:7c1b63c603788468b7679731f179d0bf
227.4 MB Preview Download
md5:6c3449f315da02cea15a8901503b93ee
201.0 MB Preview Download

Additional details

Funding

European Commission
OEMC - Open-Earth-Monitor Cyberinfrastructure 101059548
European Commission
AI4SoilHealth - AI4SoilHealth: Accelerating collection and use of soil health information using AI technology to support the Soil Deal for Europe and EU Soil Observatory 101086179

References

  • Potapov, P., Hansen, M. C., Kommareddy, I., Kommareddy, A., Turubanova, S., Pickens, A., . . . Ying, Q. (2020). Landsat analysis ready data for global land cover and land cover change mapping. Remote Sensing, 12(3), 426
  • Lang, N., Jetz, W., Schindler, K., & Wegner, J. D. (2023). A high-resolution canopy height model of the earth. Nature Ecology & Evolution, 7(11), 1778–1789
  • Potapov, P., Li, X., Hernandez-Serna, A., Tyukavina, A., Hansen, M. C., Kommareddy, A., . . . others (2021). Mapping global forest canopy height through integration of gedi and landsat data. Remote Sensing of Environment, 253, 112165
  • Potapov, P., Hansen, M. C., Kommareddy, I., Kommareddy, A., Turubanova, S., Pickens, A., . . . Ying, Q. (2020). Landsat analysis ready data for global land cover and land cover change mapping. Remote Sensing, 12(3), 426
  • Che, Y., Li, X., Liu, X., Wang, Y., Liao, W., Zheng, X., . . . others (2024). 3d-globfp: The first global three-dimensional building footprint dataset. Earth System Science Data Discussions, 2024, 1–28.
  • MacFerrin, M., Amante, C., Carignan, K., Love, M., & Lim, E. (2024). The earth topography 2022 (etopo 2022) global dem dataset. Earth System Science Data Discussions, 2024, 1–24
  • Tadono, T., Ishida, H., Oda, F., Naito, S., Minakawa, K., & Iwamoto, H. (2014). Precise global dem generation by alos prism. ISPRS annals of the photogrammetry, remote sensing and spatial1016 information sciences, 2, 71–76.
  • Takaku, J., Tadono, T., & Tsutsui, K. (2014). Generation of high resolution global dsm from alos prism. The International Archives of the Photogrammetry, Remote Sensing and Spatial Information1019 Sciences, 40, 243–248.