Estimating variable acreage of cultivated paddy fields from preceding precipitation in a tropical watershed utilizing Landsat TM/ETM

Abstract

Given a suitable high and constant temperature regime, rice can be cultivated year-round in the tropics as long as water is available. Accordingly, multiple cropping is possible in such regions. The number of croppings per year varies depending on water availability and thus annual gross cultivated acreages are subject to fluctuation. The predominant causes of this fluctuation are related to seasonal and inter-annual variability in the volume of precipitation along with the hydrological characteristics of the watershed. We analyzed the effect of preceding precipitation on cultivated paddy field acreage in the Cidanau watershed, West Java, Indonesia. The analyses had two steps; first, the manipulation of eight Landsat images of different years to identify paddy fields cultivated in the entire paddy field area. There was difficulty in identifying cultivated paddy fields because of the coexistence of various growth stages of rice; however, early growth stage paddy fields were successfully distinguished from uncultivated fields using middle infrared band values due to surface reflectance of flooded water, while more advanced growth stage fields were identified with Normalized Difference Vegetation Index values. The second step was implementation of a linear regression model to the cultivated paddy field ratio with the cumulative volume of preceding rainfall as a single independent variable. The result confirmed that there was a strong correlation between 90-day cumulative rainfall and cultivated paddy field acreage as identified by satellite images, where “90-days” implies the lifespan of paddy rice after transplantation. However, the sensitivity of cultivated paddy fields to the preceding rainfall ratio varied among sub-watersheds due to variations of hydrological characteristics and the storage capacity of the particular water source.