Abstract
Random Forest is a combination classification model based on classification and regression tree. RF model can deal with nonlinear problem, therefore, it is able to predict drought. In this paper, RF model is used to forecast drought on the Huaihe River Basin where more and more serious droughts occur. The runoff of twenty-one stations can reflect the overall drought level on Huaihe River. First, the drought has been divided into three levels depending on the SPI criteria. Then, the most important thirty factors are selected as model variables according to the Incnodepurity index after selecting preliminary screening sets. Furthermore, analysis of the drought level of the hydrological stations during the period from 1963 to 2013 is carried out with RF model. The result shows the average accuracy rate of prediction is 73%. The RF drought model is demonstrated as a new effective drought prediction model to reduce or eliminate the loss caused by drought.
Export citation and abstract BibTeX RIS
Content from this work may be used under the terms of the Creative Commons Attribution 3.0 licence. Any further distribution of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI.