Simulation of Extreme Dry and Wet Spells in Brahmaputra Basin Using K-Nearest Neighbour Model-Juniper Publishers
Juniper Publishers- Open Access Journal of Environmental Sciences & Natural Resources Simulation of Extreme Dry and Wet Spells in Brahmaputra Basin Using K-Nearest Neighbour Model Authored by Mohammed Sharif Abstract This paper describes the application of a K-nearest neighbour (KNN) weather-generating model that allows re sampling with perturbation of the observed data to simulate a) Observed maximum temperature, minimum temperature and precipitation; b) Duration of extreme wet spells; and c) Duration of extreme dry spells at ten climate stations in the Brahmaputra River Basin. The duration of extreme dry and wet spells were extracted from the 800 years of daily precipitation simulated by the K-NN model. The results of the simulation of extreme dry and wet spells carried out by the K-NN model clearly indicated that the model generated unprecedented extreme events that were not seen in the observed record. Such events are a critical input to flood management and crop productio