Research Article

Earth Science Informatics

, Volume 5, Issue 3, pp 167-179

Sobol’ sensitivity analysis of parameters in the common land model for simulation of water and energy fluxes

  • Jianhui XuAffiliated withState Key Laboratory of Information Engineering in Surveying, Mapping and Remote Sensing, Wuhan University Email author 
  • , Hong ShuAffiliated withState Key Laboratory of Information Engineering in Surveying, Mapping and Remote Sensing, Wuhan University
  • , Hongnan JiangAffiliated withResource and Environmental Science College, Xinjiang University
  • , Lin DongAffiliated withState Key Laboratory of Information Engineering in Surveying, Mapping and Remote Sensing, Wuhan University

Rent the article at a discount

Rent now

* Final gross prices may vary according to local VAT.

Get Access

Abstract

The water and energy transfer of land surface is complex due to its large spatial and temporal variability. The modeling and simulation is an important means to study land water and energy transfer, but most selection and analysis of model parameters are empirical and qualitative. This paper has proposed a method of quantitatively identifying the most influential parameters of Common Land Model through Sobol’ sensitivity analysis. Considering sensible heat flux as the model output, the first order and total sensitivity indices of 25 model input parameters are estimated using an improved Sobol’ method. The simulated results are resampled using a bootstrapping method and the corresponding sensitivity indices are calculated. Confidence intervals for the bootstrapping sensitivity indices are estimated by using a percentile method. The results show that the parameters phi0 and porsl are the most important parameters, followed by ref(2,1), tran(2,1) and bsw. Five out of 25 parameters need to have an accurate evaluation, while the other parameters are fixed to a certain value. The sensitivity indices of parameters phi0 and porsl are decreasing after precipitation, while the sensitivity indices of parameters tran(2, 1) and ref(2, 1) are increasing after precipitation.

Keywords

Common land model Sensible heat flux Sobol’ sensitivity analysis Monte Carlo Bootstrapping