A Statistical Approach to extend Fine Scale Weather Data over the Northwest Himalayas.

Abstract:

Western Himalayas is a highly complex region both in terms of topography and weather. Several factors such as elevation, slope angle and orography influence the weather of the region. Hence, accurate estimation of weather variables such as precipitation and temperature for such a complex region is difficult. Installation of rain gauges become unreasonable while excessive data handling and storage issues make climate models very expensive. To compensate this expense, meteorologists utilize climate models such as the Weather research and Forecasting (WRF) to cover larger regions with coarse spatial grids while only smaller regions are covered by fine spatial grids. This is primarily because generating data at fine spatial resolution for longer durations is computationally unaffordable. In this study, we aim to establish a statistical model which is capable of extending fine scale WRF precipitation and temperature data to a larger region. The idea is to utilize coarse and fine scale data generated by WRF for a smaller domain and employ our statistical model to predict corresponding fine scale data for larger domains. We plan to explore a total of five domains with each succeeding domain larger than the previous one. The smallest domain acts as the input domain consisting of available coarse and fine scale WRF data for precipitation and temperature. All other domains act as output domains of fine scale data. Predicted data for each larger domain is subsequently verified with the corresponding reference dataset to assess the accuracy of approach. Our study signifies three major advantages: 1) a reduced CPU run-time hour for generating fine scale weather data at high temporal resolution, 2) availability of continuous fine scale weather data for a highly complex region such as the Northwest Himalayas and 3) potential to utilize the generated data as input to other climate models to carry out researches regarding Himalayan hydroclimate, for example avalanche forecasting.