Interpretation of Weather Forecasts for Tornado and Cloudburst using Data Mining Techniques

Abstract

With the advent of digital computers and their continuous increasing processing power, newlinethe Numerical Weather Prediction (NWP) models which solve a close set of equations newlinerepresenting atmospheric flow, have been adopted by most of the meteorological services to newlineissue day to day weather forecasts. These forecasts are issued for public in general, farmers, newlinepilots, water works managers, health departments, planners, disaster management services etc. newlineNWP models continue to improve in resolution (both horizontal and vertical) as well as newlinesophistication in including various atmospheric processes. Despite this, there are various newlinelimitations, specifically: newline(i) There are many important processes and scales of motion in the atmosphere especially newlineSub-Grid scale weather phenomenon that cannot be explicitly resolved with present newlinemodels. Some of the significant Sub-Grid scale phenomenons are Tornado and newlinecloudbursts. Various empirical techniques are used by the experienced forecasters to newlineinfer these events. newline(ii) The NWP output consists of mainly flow patterns namely wind, temperature, humidity, newlineand pressure fields at various temporal and spatial levels. The forecast of actual weather newlineelements like rain/ snow etc. are derived from the NWP output products through newlinestatistical relationship, known as Model Output Statistics (MOS). But MOS is not a newlinetheoretically stable process as it requires longer datasets on time scale to derive the MOS newlinerelationships for forecasting of weather elements. Longer and consistent datasets are not newlineavailable because of frequent revisions of NWP model. newlineThere is thus a strong need for searching alternative tools to MOS for interpretation of weather newlinepatterns provided by NWP models into weather elements including cloudburst occurrence and newlinetornado. According to Literature Survey, Intelligent systems have been applied generally for newlineprocesses like decision making applications of evacuation of public in case of cyclone hit, newlineprobability of occurrence of rainfall / snow etc. newline