A minimal soil moisture model fit to environmental data from multiple pasture locations in Taranaki, New Zealand

Abstract

The application of models and precision technology to optimize productivity and sustainability is increasingly common in agriculture. Soil moisture (SM) modelling is an important component of pasture growth modelling, or for runoff/catchment modelling. This paper examines a minimal modelling approach for SM modelling using soil moisture data from 10 locations in the Taranaki region, New Zealand. Several simple compartment-based models are tested with/without daylight and temperature effects on terms relating to SM loss, and gain from rainfall. It was found that SM dynamics differed from site to site, with a simple loss dynamic proportional to current SM level dominating in soils with moisture levels above field capacity. A SM loss term modified by temperature and day length described sites where SM was below field capacity. Thus, a simple model was able to fit locational data, and distinguished between differing dynamics based on SM level relative to field capacity. This is model is a first step towards a model-predictive approach to soil moisture modelling requiring a minimum of measured inputs.