Regionalizing Root‐Zone Soil Moisture Estimates From ESA CCI Soil Water Index Using Machine Learning and Information on Soil, Vegetation, and Climate
The European Space Agency (ESA), through the Climate Change Initiative (CCI), is currently providing nearly 4 decades of global satellite‐observed, fully homogenized soil moisture data for the uppermost 2–5 cm of the soil layer. These data are valuable as they comprise one of the most complete remot...
Gespeichert in:
Veröffentlicht in: | Water resources research 2021-05, Vol.57 (5), p.n/a |
---|---|
Hauptverfasser: | , , , , |
Format: | Artikel |
Sprache: | eng |
Schlagworte: | |
Online-Zugang: | Volltext |
Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
Zusammenfassung: | The European Space Agency (ESA), through the Climate Change Initiative (CCI), is currently providing nearly 4 decades of global satellite‐observed, fully homogenized soil moisture data for the uppermost 2–5 cm of the soil layer. These data are valuable as they comprise one of the most complete remotely sensed soil moisture data sets available in time and space. One main limitation of the ESA CCI soil moisture data set is the limited soil depth at which the moisture content is represented. In order to address this critical gap, we (a) estimate and calibrate the Soil Water Index using ESA CCI soil moisture against in situ observations from the International Soil Moisture Network and then (b) leverage machine learning techniques and physical soil, climate, and vegetation descriptors at a global scale to regionalize the calibration. We use this calibration to assess the root‐zone soil moisture for the period 2001–2018. The results are compared against the European Centre for Medium‐Range Weather Forecasts, ERA5 Land, and the Famine Early Warning Systems Network Land Data Assimilation System reanalyses soil moisture data sets, showing a good agreement, mainly over mid latitudes. This work contributes to the exploitation of ESA CCI soil moisture data, while the produced data can support large‐scale soil moisture‐related studies.
Key Points
We use machine learning to regionalize the calibration of the Soil Water Index based on soil, climate, and vegetation descriptors
The results are compared to reanalysis data sets, indicating added value to the results of the machine learning calibration |
---|---|
ISSN: | 0043-1397 1944-7973 |
DOI: | 10.1029/2020WR029249 |