Parameter sensitivity of SWAP-PEARL models for pesticide leaching in macroporous soils

Abstract

Pesticide transport simulation by SWAP-PEARL (Soil-Water-Atmosphere-Plant and Pesticide Emission Assessment at Regional and Local scales) models can help to predict pesticide leaching at regional scales. For reasons of economic and time efficiency, measurement efforts should be prioritized towards critical parameters. The objective of this research is to perform a Morris screening and Sobol-Jansen sensitivity analysis to SWAP-PEARL models, using a reasonable worst-case scenario. Three pesticide compounds were analyzed:, bentazon (zero sorption), imidacloprid (moderately sorbed), and compound I (highly sorbed). Initial macropore and pesticide parameter values were varied by +/- 20% to generate parameter ranges. The outputs analyzed were the concentration in drainage water, the average concentration in groundwater between 1 and 2 m, and the concentration in the soil system at 100-cm depth. Influential parameters found through the Morris method were analyzed using the Sobol-Jansen method. The results for bentazon indicate that the degradation half-life (DT50), the bottom depth of the internal catchment (z(ic)), and the proportion of the internal catchment at the soil surface (p(ic_0)) are critical parameters in all the outputs analyzed. For imidacloprid and compound I, the most relevant parameters for drainage output are the Freundlich sorption exponent (F-exp) and z(ic); for groundwater, the relevant parameters are F-exp, the bottom depth of static macropores (z(st)), and p(ic_0); and for soil concentrations at 100-cm depth, the relevant parameters are F-exp, z(ic), and p(ic_0). The Morris and Sobol-Jansen methods produce the same results for the first position in the ranking. Measurement efforts should be performed to update national soil databases, including critical pesticide and macropore parameters.