The usefulness of weather forecasting relies on Mine-Weather’s ability of scoring relevant weather attributes. Mine-Weather is shipped with a pre-loaded set of standard scoring algorithms used for statistical forecast verification. Mine-Weather platform encompasses numerous scoring algorithms based on statistical methods for easy verification and evaluation of weather forecasts.
Score is a measure for evaluating the quality of various forecasts. Scores show a possible correlation between forecast and observed values and the accuracy of these values. Mine-Weather’s scoring algorithm transforms all statistical measurements into numerical metrics ranging from 0 to 100. This helps a decision maker in comparing scores on each weather attribute and arrive at the most accurate forecast for a specific type of weather event. In this way decision makers can track the right forecast vendor from a specific weather parameter.
Numerous challenges are involved in verifying weather forecasts. A host of weather attributes such as numeric point forecasts (hourly/daily temperature, wind speed, etc.), some dichotomous (yes/no, hit/miss, daily precipitation, etc.) and some non-numeric (wind direction, cloud cover, etc.) are the major ingredients that constitute a weather forecast. Having a myriad of different types of weather attributes makes it a challenging exercise while comparing the forecast with actual values.
Mine-Weather proposes the following solutions to overcome the challenges:
- Numeric Variables – Scoring are based on the numeric comparison of forecasted values against the observed actuals.
- Dichotomous Attributes – Scores are calculated by comparing the forecasts against observed actuals using statistical contingency tables / by computing categorical statistics.
- Special Weather Variables – Scores are measured across one or all weather variables at short time intervals during particular event like storm or hurricane.
At present, Mine-Weather provides forecast verification for temperature, precipitation, wind speed/gust, dew point, snow fall, humidity, and wet bulb. Composite scoring is another capability the Mine-Weather platform inherits to judge the overall performance of each vendor. Scores across different weather attributes are consolidated to form a composite score that can be prioritized based on specific weather attributes. Download our WhitePaper to get an in-depth understanding of the various scoring algorithms being used in Mine-Weather.
Mine-Weather platform can be customized according to the specific needs of the client. It adopts a flexible weather forecast verification methodology that is sophisticated enough to add advanced scoring methods. Mine-Weather uses standard verification methods and calculations for performing qualitative assessment of forecasts. It can be tailor made to fit the specific forecast needs of any client requiring to probe the accuracy of weather forecasts.