Making Climate Forecasts with Python and Numpy
- Audience level:
Python and Numpy can be used to perform quick analyses on large datasets, which is especially important for weather and climate prediction and monitoring. This presentation will show how Python, Numpy, and Numpy Mask arrays were used to develop an application that produces climate forecasts using information from numerical weather models.
Climate forecasting and monitoring often requires developing software that can crunch numbers. Lots of numbers. Statistical analysis is often performed across many data points over time and space in the climate framework. This presentation will show an example of how Python and Numpy was used to develop the Forecast Consolidation Tool (FCT) at the Climate Prediction Center (CPC), an application that produces optimally combined forecasts from various numerical weather prediction models, based on their past performance. This presentation will include how Python, Numpy, and Mask arrays can be used to perform statistical calculations on large datasets, specifically applied to making climate forecasts. Forecast graphics from the FCT will be shown, as an example of how geospatial maps can be easily created using Matplotlib. Lessons learned will also be discussed regarding how Numpy Mask arrays should be used with caution, especially with handling missing data (a common issue for many science and engineering applications).