Application of MODIS-Based Monthly Evapotranspiration Rates in Runoff Modeling: A Case Study in Nebraska,USA

Jozsef Szilagyi

doi:10.4236/ojmh.2013.34021

Abstract

Daily and monthly flow-rates of the Little Nemaha River in Nebraska were simulated by the lumped-parameter Jakeman-Hornberger as well as a distributed-parameter water-balance accounting procedure for the 2003-2008 and 2000-2009 periods, respectively, with and without the help of the MODIS-based monthly estimates of evapotranspiration (ET) rates. While the daily lumped-parameter model simulation accuracy remained practically unchanged with the inclusion of the monthly MODIS-based ET rates interpolated into daily values (R2 of 0.66 vs 0.68, simulated to measured runoff ratio remaining the same 96%), the monthly water-balance accounting model outcomes did improve to some extent (from an R2 of 0.67 to 0.7 with simulated to measured runoff ratio of 72% vs 115%). In both cases the models had to be slightly modified for accommodation of the ET rates as predefined input values, not present in the original model setups. These results indicate the potential practical usefulness of satellite-derived ET estimates (CREMAP values in the present case) in monthly water-balance modeling. CREMAP is a calibration-free ET estimation method based on MODIS-derived daytime surface temperature values in combination of basic climatic variables, such as air temperature, humidity and solar radiation within a Complementary Relationship framework of evaporation.

Highlights

With the free public availability of the ~1-km, global, Moderate Resolution Spectroradiometer (MODIS, available at https://lpdaac.usgs.gov/lpdaac/products/modis_ products_table) data since 2000, the number of remotesensing based, basin-scale evapotranspiration (ET) estimation methods have seen an unprecedented growth
While the daily lumped-parameter model simulation accuracy remained practically unchanged with the inclusion of the monthly MODIS-based ET rates interpolated into daily values (R2 of 0.66 vs 0.68, simulated to measured runoff ratio remaining the same 96%), the monthly water-balance accounting model outcomes did improve to some extent
These results indicate the potential practical usefulness of satellite-derived ET estimates (CREMAP values in the present case) in monthly water-balance modeling

Summary

Introduction

With the free public availability of the ~1-km, global, Moderate Resolution Spectroradiometer (MODIS, available at https://lpdaac.usgs.gov/lpdaac/products/modis_ products_table) data since 2000, the number of remotesensing based, basin-scale evapotranspiration (ET) estimation methods have seen an unprecedented growth. The different approaches are based on the application of a vegetation index and/or the land surface temperature to solve the energy balance equation at the ground Common to all these methods is that they are based on simplifying assumptions and require some sort of parameter calibration, typically aided by precipitation and runoff data at the basin scale. While the CREMAP ET rates have proven valuable for specifying spatially varying regional recharge rates in Nebraska and Hungary [9,10,11], as well as defining net recharge to the groundwater as a function of vadose-zone depth in the shallow groundwater area of the Platte River in Nebraska [12], their practical value in runoff modeling has not been investigated, motivating the present study. In the present study the state-wide monthly CREMAP ET rates of Szilagyi [3] are utilized for the Little Nemaha watershed in the south-eastern part of Nebraska, USA (Figure 1)

Study Site and Runoff Model Descriptions

Results and Discussion

Conclusion