Estimating temperature of mulched and bare soil from meteorological data

Abstract

In order to investigate the application potential for soil solarization in the southern US without conducting labor intensive field tests and expensive experiments, a numerical model has been developed to estimate the temperature profile of both mulched and bare soils. Atmospheric and soil conditions, as well as the transmissivity, reflectivity and emissivity of mulch are considered in the model. The required dynamic inputs are hourly measurements of global radiation, air temperature, dewpoint, wind speed and rainfall. The model was validated using hourly observations from 12 contiguous days of July 6–18, 1990 at the North Carolina State University Horticultural Crops Research Station near Clinton. Different weather occurred during the period. The model worked very well on both clear and rainy days except July 17 when large, rapid changes of the air temperature and solar radiation occurred. However, the percentages of the absolute differences less than 2.0°C between the hourly estimated and measured soil temperatures at 10, 20, and 30 cm were 89, 95 and 95 for mulched soil, and 94, 98 and 100 for bare soil, respectively. The correlation between estimated and measured temperatures yielded R-square values between 0.82 and 0.93. The model was very successful to satisfy the main objectives in this study. Model sensitivities to 23 parameters were analyzed. Relative sensitivity coefficients were higher for soil bulk density, quartz fraction, and mulch transmissivity to solar radiation, than for surface roughness length, soil clay fraction and mulch transmissivity to long wave radiation.