Intelligent retrieval of historical meteorological data

Abstract

Large archives of historical meteorological data offer novel opportunities for weather forecasting and meteorological research. However, effective exploitation of such archives is impeded by the logistics of managing large volumes of data and the difficulty of identifying relevant data. This thesis presents a system called MetVUW Workbench that addresses these problems. MetVUW Workbench is a case-based retrieval system designed to assist weather forecasters and meteorology researchers. Users employ a graphical query language to compose queries in terms of high-level meteorological phenomena. The system retrieves past weather situations (cases) that are good matches to the query. Cases are stored using a relational database and are retrieved in two stages. The first stage quickly identifies cases that are likely to be good matches to the query. The second stage computes a graded partial match between the query and the cases identified by the first stage. A prototype has been implemented and tested with a case base of 2500 cases. Evaluations of the prototype indicate that the design should scale to case bases several times larger.