Using a groundwater adjusted water balance approach and copulas to evaluate spatial patterns and dependence structures in remote sensing derived evapotranspiration products

Mohsen Soltani, Julian Koch, Simon Stisen

Research output: Contribution to journalArticleResearchpeer-review

6 Citations (Scopus)

Abstract

This study aims to improve the standard water balance evapotranspiration (WB ET) esti-mate, which is typically used as benchmark data for catchment-scale ET estimation, by accounting for net intercatchment groundwater flow in the ET calculation. Using the modified WB ET approach, we examine errors and shortcomings associated with the long-term annual mean (2002–2014) spatial patterns of three remote-sensing (RS) MODIS-based ET products from MODIS16, PML_V2, and TSEB algorithms at 1 km spatial resolution over Denmark, as a test case for small-scale, energy-limited regions. Our results indicate that the novel approach of adding groundwater net in water balance ET calculation results in a more trustworthy ET spatial pattern. This is especially relevant for smaller catchments where groundwater net can be a significant component of the catchment water balance. Nevertheless, large discrepancies are observed both amongst RS ET datasets and compared to modified water balance ET spatial pattern at the national scale; however, catchment-scale analysis highlights that difference in RS ET and WB ET decreases with increasing catchment size and that 90%, 87%, and 93% of all catchments have ∆ET < ±150 mm/year for MODIS16, PML_V2, and TSEB, respectively. In addition, Copula approach captures a nonlinear structure of the joint relationship with multiple densities amongst the RS/WB ET products, showing a complex dependence structure (correlation); however, among the three RS ET datasets, MODIS16 ET shows a closer spatial pattern to the modified WB ET, as identified by a principal component analysis also. This study will help improve the water balance approach by the addition of groundwater net in the ET estimation and contribute to better understand the true correlations amongst RS/WB ET products especially over energy-limited environments.

Original languageEnglish
Article number853
Number of pages25
JournalRemote Sensing
Volume13
Issue number5
DOIs
Publication statusPublished - 1 Mar 2021

Keywords

  • Bivariate density estimation and similarity assessment
  • Energy-limited environment
  • Evapotranspiration
  • Remote-sensing products
  • Spatial pattern evaluation
  • Water balance approach
  • DK-model

Programme Area

  • Programme Area 2: Water Resources

Fingerprint

Dive into the research topics of 'Using a groundwater adjusted water balance approach and copulas to evaluate spatial patterns and dependence structures in remote sensing derived evapotranspiration products'. Together they form a unique fingerprint.

Cite this