Search Filter

Keywords:

 

Asian Journal of Agriculture and Development (AJAD) - Call for papers!

Runoff Prediction Model Based on Soil Moisture Analysis

(Indonesia), Doctor of Philosophy in Forest Resources Management (University of the Philippines Los Baños)

Dissertation Abstract:

 

Based on the resulting seed quality and land equivalent ratio (LER), the best intercropping proportions were one row of corn with four rows of soybean and two rows of corn with three rows of soybean. These can also be used as alternative method for seed production of corn and soybean, particularly for planting purposes.

A study of Padas and Wader watersheds located in Upper Solo watershed, Central Java, Indonesia showed that the area was characterized by two distinct seasons - dry season from May to October and wet season from November to April.

A deterministic, event, and lumped model was used to predict runoff from the two watersheds. The model consists of four parameters, namely: 1) maximum rate of seepage from the soil water zone in inches per day (S); 2) maximum soil moisture loss readily available for evapotranspiration in inches (C); 3) fraction of seepage from the soil water zone that becomes runoff(F); and 4) maximum infiltration rate in inches per hour (Fmax).

These parameters were optimized by minimizing the sum of squares of the difference between observed and simulated runoff. The model required as input the daily rainfall and the estimated potential evapotranspiration, which was computed using the Penman method.

Streamflow records for 10 years ( 1976-1977 and 1979-1986) of Padas watershed (3,485 ha) and 1 years ( 1976-1977 and 1978-1986) of Wader subwatershed (202 ha) were used to test the model's validity. Data of monthly flows for four years were used to estimate the optimal values of the parameters.

The final test of the model's adequacy was in a comparison of observed and simulated runoff, which showed that the observed and simulated runoff values were not significantly different. This was based on the results obtained from four statistical measures used to test the model. The model did a better simulation in Wader (the smaller watershed) than in Padas (the larger one).