Premium
Topographic Wetness Index as a Proxy for Soil Moisture: The Importance of Flow‐Routing Algorithm and Grid Resolution
Author(s) -
Riihimäki H.,
Kemppinen J.,
Kopecký M.,
Luoto M.
Publication year - 2021
Publication title -
water resources research
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.863
H-Index - 217
eISSN - 1944-7973
pISSN - 0043-1397
DOI - 10.1029/2021wr029871
Subject(s) - topographic wetness index , digital elevation model , algorithm , flow routing , routing (electronic design automation) , environmental science , water content , grid , proxy (statistics) , soil science , remote sensing , data mining , hydrology (agriculture) , computer science , geology , machine learning , geotechnical engineering , geodesy , computer network
The Topographic Wetness Index (TWI) is a commonly used proxy for soil moisture. The predictive capability of TWI is influenced by the flow‐routing algorithm and the resolution of the Digital Elevation Model (DEM) that TWI is derived from. Here, we examine the predictive capability of TWI using 11 flow‐routing algorithms at DEM resolutions 1–30 m. We analyze the relationship between TWI and field‐quantified soil moisture using statistical modeling methods and 5,200 study plots with over 46 000 soil moisture measurements. In addition, we test the sensitivity of the flow‐routing algorithms against vertical height errors in DEM at different resolutions. The results reveal that the overall predictive capability of TWI was modest. The highest r 2 (23.7%) was reached using a multiple‐flow‐direction algorithm at 2 m resolution. In addition, the test of sensitivity against height errors revealed that the multiple‐flow‐direction algorithms were also more robust against DEM errors than single‐flow‐direction algorithms. The results provide field‐evidence indicating that at its best TWI is a modest proxy for soil moisture and its predictive capability is influenced by the flow‐routing algorithm and DEM resolution. Thus, we encourage careful evaluation of algorithms and resolutions when using TWI as a proxy for soil moisture.