Comparing Crack Volume Estimations On a Vertisol Monitored in South Texas for Ten Years.

Water distribution into infiltration and runoff is a very dynamic process in landscapes with shrinking-swelling soils due to surface cracking. Estimates of soil crack volume can aid in simulating water infiltration and redistribution in cracking soils. To estimate crack volume, commonly used approaches are based on the soil coefficient of linear extensibility (COLE) and shrinkage characteristic curve (Shrinkage Curve). The objectives of this work were to 1. Estimate vertical crack volume using measurements of crack dimensions; 2. Compare COLE-based and Shrinkage Curve-based estimates of crack volume to estimates derived from field measurements. Crack dimensions were monitored under short prairie vegetation at a 100 m2 site of a Laewest clay with gilgai in Victoria County, Texas, on 42 dates during 1989-1998. Location of cracks was recorded on scaled diagrams and crack width was color coded with limits of 0.5, 1, 2, 5 and 7 cm. For the largest cracks, crack depth and exact width were recorded at certain locations. Gravimetric soil water content was measured for 10, 25, 50, 75 and 100 cm depths on microhighs and microlows adjacent to the study site on 50 dates; 18 dates simultaneously with crack measurements and was converted to volumetric values using soil horizon based bulk densities. Crack length was calculated for microhighs, microslopes, and microlows. Crack volume per unit area was estimated from crack lengths, widths, and measured or estimated crack depths. For cracks where depth was not measured, an empirical relationship between crack width and total crack depth was used. Crack volume was calculated assuming an equi-lateral triangular cross section of vertical cracks. The two crack volume estimations (COLE and Shrinkage Curve) will be evaluated in comparison to crack volume derived from field measurements in terms of R2, RMSD, and p-value, and will be presented for microhighs and microlows.