<![CDATA[Drought is often categorized as a natural disaster, which receives less attention compared to flood, earthquake, and landslide due to its slow development characteristic. However, the loss due to drought events especially for farmers is comparable with other hazards. To reduce the impact of droughts, a drought early warning system with a lead time of several months ahead is needed. This paper, therefore, exploits the insight of drought early warning system that can be developed in Indonesia. Results show that drought analysis to calculate the forecasted drought severity can be separated into two algorithms. First algorithm is developed to calculate the distribution parameters or CDF values for each month and in each grid cell using historical data that will be used in second algorithm. The second algorithm is built to calculate the drought severity index using forecast data and distribution parameters or CDF values. Outcomes of the analysis are forecasted drought regions and drought index in each province in Indonesia, showing the top and bottom boundaries of the forecast essembled models. Bias correction must be carried out if the hindcast data are available in order to obtain more accurate drought forecasting results. The insight of drought early warning system that is described in this paper hopefully can guide the researchers to develop the drought early warning system in Indonesia.]]>
Copyrights © 2017
