Models for spatio-temporal prediction are needed that account for mass transport processes. Past work has concentrated on water erosion but has neglected water deposition and soil transport by tillage. Understanding these processes is essential to successful numerical modeling and for the placement of sites to monitor SOC changes for carbon accounting. To better understand and model these processes, two high-resolution (25 m) soil sample grids were collected on separate agricultural fields (chisel ploughed, corn soybean rotation). SOC and Cs-137 (to model erosion/deposition) were measured for each soil sample point. Five meter resolution data were collected with GPS to provide topographic information (DEM). A unique characteristic of the study fields is that they contain small (100 m from bottom to top, 5 meters relief) which may be closed sedimentary systems. For such watersheds, the mass balance of soil transport should sum to zero. Mass balance was calculated for several watersheds near Ames, Iowa. One watershed was found to have significant sediment input from outside sources. The other watersheds had mass balances within potential losses/ gains from wind erosion (±2.5-5 Mt/ha/year). Erosion/deposition models based on Cs-137 concentration (Walling and He method) are giving reasonable results and do not show systematic bias towards erosion or deposition. These results demonstrate the need to take into account landscape redistribution of carbon when using the benchmark approach to assess temporal changes in SOC inventory.