|De mello, Carlos|
|De mello, Jose|
Submitted to: ASABE Annual International Meeting
Publication Type: Proceedings
Publication Acceptance Date: 5/30/2009
Publication Date: 6/21/2009
Citation: De Mello, C.R., Avila, L.F., Norton, L.D., De Mello, J.M. 2009. Soil Moisture Mapping at a Watershed of the Mantiqueira Range, MG, Brazil [abstract]. American Society of Agricultural and Biological Engineers Annual International Meeting. Paper No. 095566. Interpretive Summary: Water content of soil in a watershed is important to know in order to predict when and how much runoff and erosion will occur. We conducted a study to measure soil water content through the seasons at different locations in a headwater watershed that feeds into a hydroelectric facility over one year. The data collected were subjected to various statistical procedures to accurately map the soil water content across the watershed. It was found that the slope of the land and type of land use were most important in explaining the different water contents measured over the various seasons. There were large differences in soil water content noted between grasslands and the native Atlantic forest over the seasons with the Atlantic forest conserving more soil water throughout the year. The implications of this research is that limited data on soil moisture content and a map of land slope and use can be combined to more accurately predict the runoff from a watershed and estimate water flows and sediment loadings into downstream hydroelectric production facilities.
Technical Abstract: Soil moisture is an important hydrological variable that characterizes the soil water dynamics influencing surface runoff generation and consequently sediment transport. This work aimed to analyze the spatial patterns of surface soil moisture, to identify the elements that exert the most influence in the context of its spatial distribution on surface soil moisture through time in an experimental catchment located in Mantiqueira Range region, southeast Brazil. The measurements of surface soil moisture were carried out between May/2007 and May/2008, every 15 days, on the top 20 cm of soil profile by TDR equipment, in 69 points. Geostatistical procedures were applied in all steps of work. Firstly, the spatial continuity of each data set for surface soil moisture was evaluated, modeling the experimental semi-variogram by testing of Exponential, Spherical and Gaussian models which were fitted applying the Weighted Minimum Square (WMS), Ordinary Minimum Square (OMS) and Maximum Likelihood (ML) methods. We verified good performance for all models, and especially, those fitted by WMS. However, the exponential model was found to be the best. For development of surface soil moisture mapping co-kriging using the land slope as a secondary variable was applied. Important influence of slope was verified as important during the dry season. Rainfall regime controls the surface soil moisture in the wet season due to the high moisture contents. Land-use was another local fundamental factor identified as important, especially under dry conditions. In the Atlantic Forest, the surface soil moisture, over the period, decreased slower than the grassland sites which showed a much faster reduction.