Computationally inexpensive identification of noninformative model parameters by sequential screening

Abstract

Environmental models tend to require increasing computational time and resources as physical process descriptions are improved or new descriptions are incorporated. Many-query applications such as sensitivity analysis or model calibration usually require a large number of model evaluations leading to high computational demand. This often limits the feasibility of rigorous analyses. Here we present a fully automated sequential screening method that selects only informative parameters for a given model output. The method requires a number of model evaluations that is approximately 10 times the number of model parameters. It was tested using the mesoscale hydrologic model mHM in three hydrologically unique European river catchments. It identified around 20 informative parameters out of 52, with different informative parameters in each catchment. The screening method was evaluated with subsequent analyses using all 52 as well as only the informative parameters. Subsequent Sobol's global sensitivity analysis led to almost identical results yet required 40% fewer model evaluations after screening. mHM was calibrated with all and with only informative parameters in the three catchments. Model performances for daily discharge were equally high in both cases with Nash-Sutcliffe efficiencies above 0.82. Calibration using only the informative parameters needed just one third of the number of model evaluations. The universality of the sequential screening method was demonstrated using several general test functions from the literature. We therefore recommend the use of the computationally inexpensive sequential screening method prior to rigorous analyses on complex environmental models.

Toward seamless hydrologic predictions across spatial scales

“…We selected soil porosity as an example to visualize existing shortcomings because it is one of the most common parameters in many LSMs/HMs. This parameter controls the dynamic of several state variables and fluxes such as soil moisture, latent heat, and soil temperature, and its sensitivity has been demonstrated in various studies (Goehler et al, 2013;Cuntz et al, 2015;Mendoza et al, 2015;Cuntz et al, 2016). A representation of the porosity of the top 2 m soil column in these models over the Pan-European domain (Pan-EU) is shown in Fig.…”

“…Determine a set of the most sensitive model parameters through a sensitivity analysis (SA). For computationally expensive LSMs such as CLM or Noah-MP, computationally frugal methods such as the elementary effects method (Morris, 1991), its enhanced version such as that proposed by Cuntz et al (2015), or the distributed evaluation of local sensitivity analysis (DELSA; Rakovec et al, 2014;Mendoza et al, 2015) are of particular interest because use of the popular standard Sobol' method (Sobol', 2001) can be computationally expensive although still possible .…”

Regional analysis of parameter sensitivity for simulation of streamflow and hydrological fingerprints

“…The different number of dominant parameters might also be due to correlated mHM parameters which we sorted out before sensitivity analysis. In contrast, Cuntz et al (2015) considered the degree of correlation between mHM parameters as rather minor to be interfering with parameter identification.…”

“…The mHM is conceptualised on the basis of grid cells, and has been applied to a wide range of mesoscale river catchments (10 1 -10 4 km 2 ; Kumar et al, 2010;Samaniego et al, 2010aSamaniego et al, , 2011Cuntz et al, 2015;Rakovec et al, 2016). Gridded information is implemented in mHM at three levels: morphology (level 0), hydrology (level 1), meteorology (level 2), with l 0 l 1 ≤ l 2 denoting the relative sizes of the grid cells at the respective data level (Kumar et al, 2010).…”

“…In this context, Herman et al (2013) showed that the longterm water balance is dominated by only very few parameters, irrespective of the hydrological conditions and of the model. Cuntz et al (2015) performed a global Sobol sensitivity analysis on the hydrologic model mHM. For three distinct humid and arid European catchments this analysis always resulted in about 20 informative parameters, though the dominant parameter sets were composed very differently.…”