Premium
Estimating detailed soil water profile records from point measurements
Author(s) -
FernándezGálvez J.,
Simmonds L. P.,
Barahona E.
Publication year - 2006
Publication title -
european journal of soil science
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.244
H-Index - 111
eISSN - 1365-2389
pISSN - 1351-0754
DOI - 10.1111/j.1365-2389.2005.00761.x
Subject(s) - water content , infiltration (hvac) , environmental science , soil science , soil water , soil horizon , hydrology (agriculture) , sampling (signal processing) , groundwater recharge , interpolation (computer graphics) , geology , groundwater , meteorology , geotechnical engineering , geography , filter (signal processing) , computer science , aquifer , computer vision , animation , computer graphics (images)
Summary Temporal and spatial patterns of soil water content affect many soil processes including evaporation, infiltration, ground water recharge, erosion and vegetation distribution. This paper describes the analysis of a soil moisture dataset comprising a combination of continuous time series of measurements at a few depths and locations, and occasional roving measurements at a large number of depths and locations. The objectives of the paper are: (i) to develop a technique for combining continuous measurements of soil water contents at a limited number of depths within a soil profile with occasional measurements at a large number of depths, to enable accurate estimation of the soil moisture vertical pattern and the integrated profile water content; and (ii) to estimate time series of soil moisture content at locations where there are just occasional soil water measurements available and some continuous records from nearby locations. The vertical interpolation technique presented here can strongly reduce errors in the estimation of profile soil water and its changes with time. On the other hand, the temporal interpolation technique is tested for different sampling strategies in space and time, and the errors generated in each case are compared.