Performance of soil moisture balance computer simulation models in predicting soil moisture balance under different tillage methods in semi-arid areas of Morogoro, Tanzania
Loading...
Date
1999
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Sokoine University of Agriculture
Abstract
Soil moisture balance computer simulation models play a very important role in providing scientists with a tool able to rapidly calculate the magnitude of various processes affecting the field water balance on any soil under any given climatic and vegetation conditions. This study evaluates the performance of three soil water balance computer simulation models on different soil water conservation tillage plots. The three models are: SWAM (Soil Water Availability Model), a sub-model of PARCH (Bradley and Crout, 1994), PATHSMM (PARCHED THIRST soil moisture model), a sub-model of PARCHED-THIRST (Young and Gowing. 1996) and BUDGET (Raes, 1996) a sub-model of IRSIS (Irrigation Scheduling Information System) (Raes et al., 1988). Experiments were conducted at Sokoine University of Agriculture (SUA) farm, Morogoro, Tanzania, on a sandy clay soil. Maize crop (zea mays L) was planted for two consecutive rainy seasons namely; short rainy season and long rainy season. The soil tillage treatments were: flat cultivation (FC), zero tillage (ZT), staggered ridges (SR) contour ridges (CR) and tied-ridges (TR). All the models used in the study had better soil moisture balance simulations on the flat cultivation (FC) treatment than other treatments. However, the models tended to overpredict the soil moisture content. The performance of the PATHSMM model was very remarkable on all the tillage treatments over the two rainy seasons. There was good correlation between simulated and actual soil moisture content. Correlation coefficients ranged between 0.64 - 0.95 for PATHSMM simulations. From the study it is recommended that:
(a) The BUDGET model be improved by incorporating tillage, macropore flow and crusting factors in it.
(b) The PATHSMM model be tested on other soil types and treatments. The model needs to develop a wide base in terms of tillage options.
(c) All the three models should develop the ability to simulate soil profile moisture changes due to changes in soil bulk density.
(d) More research needs to be done to assess the models’ performance under very dry conditions.
Description
Master's Dissertation
Keywords
Soil moisture, Soil computer simulation models, Predicting soil moisture, Semi-arid areas, Morogoro, Tanzania