Estimating Antecedent Soil Moisture Effect on Surface Cracking of a Vertisol.

Watershed models that include soil cracking modules use current soil moisture and soil shrinkage relationship to predict cracking extent. Extreme cracking events, measured in field conditions in 1997-1998, indicate that weather-related soil moisture storage prior to cracking may have a significant effect on surface crack area density. The objective of this study was to quantify the effect of antecedent soil moisture conditions on the surface crack area density. Ten years of temporal and spatial variability of a cracking Vertisol was measured originally to answer Taxonomy related questions. Cracking on a 10 x 10-m plot of a Laewest clay (fine, smectitic, hyperthermic Typic Hapludert) with native tallgrass vegetation was measured and plotted to scale on 42 dates within 1989-1998. Gravimetric soil moisture was determined on 50 dates, from which 18 dates corresponded with crack measurements. Gilgai microtopograhy was mapped from elevation measurements. Crack area density was calculated from the diagrams. In this study, crack area density on microhighs was analyzed in relation to prior weather conditions. From daily precipitation, temperature, solar radiation, wind speed, and relative humidity, reference evapotraspiration and a water-availability index were calculated. Drying, uniform and wetting soil moisture profile conditions were assessed by comparing soil moisture content at 10 cm to 25 cm of the soil profile. The results show that recharging maximum of the soil characterized by the water-availability index and antecedent moisture prior to cracking are curvilinearly related (r2=0.5). The overall 10-year variation of crack area density, including the extreme cracking events, can be explained up to 92 % by models including current and antecedent soil moisture. Prediction accuracy varies on drying-wetting conditions.