Multiyear Seasonal Precipitation, Temperature Assessment and Forecasting using ARIMA Forecast Model: A Case Study of Upper Dudh Koshi Basin, Nepal

Climate warming has been a major concern in the glacierized catchments upper Dudh Koshi River basin, Eastern Himalayan region in Nepal. Complex topography and sparse hydro-meteorological measurements result to limited hydroclimatic data that affects accuracy in simulation of forecasting models. In addition, climate change is impacting mountain water resources for drinking, agriculture and hydropower, and due to increased glacier and ice loss, they are pressing issue in the region. The goal of this study is to investigate if seasonal and multiyear forecast of climatic condition is possible. We used Autoregressive Integrated Moving Average (ARIMA) and multiplicative Seasonal Autoregressive Integrated Moving Average (SARIMA) to forecast the temperature and precipitation of Himalayan basin. Monthly data from four stations (4 precipitation and 1 temperature stations) are considered from 2015 to 2021 to evaluate the seasonal forecast of temperature and precipitation. Multiplicative ARIMA model is validated with the observed data for time period 2015-2021. Since model produced good results, we further used to predict the climatic condition for period from 2020 to 2050. We present the ARIMA forecast model results comparing with CIMP5 future projection data considering different scenarios to determine reasonable accuracy and model bias. Our findings further evaluate the applicability and scalability of ARIMA model for predicting climatic condition and variability to address the assessment of climate change challenges in data scare high altitude regions.