climate-predictability-tool

What problem does it solve?

This Skill facilitates the development of statistical/empirical seasonal climate forecasting using multiple linear regression, enabling the creation of models that predict climate variables based on historical data.

Core Features & Use Cases

• Regression Modeling: Offers CCA and PCR regression models for climate forecasting.
• Model Validation: Uses cross-validation to assess model skill and select the best model.
• Probabilistic Forecasting: Creates probabilistic forecasts based on regression model errors.
• Use Case: Imagine you want to predict rainfall in West Africa for the upcoming season. Use this Skill to build a regression model using historical GCM forecasts and observations, validate its skill, and generate probabilistic forecasts.

Quick Start

Use the climate-predictability-tool skill to build a regression model using GCM forecasts and observations from the IRI Data Library for West African precipitation.