Date of this Version



This practical research report was submitted to the faculty of Purdue University in partial fulfillment of the requirements for the Geodata Science for Professionals Master of Science degree.


Atmospheric rivers (AR) are long and narrow filaments in the atmosphere that transport water vapor in the lower troposphere. They release water vapor in the form of rain or snow when they make landfall and can therefore be linked to floods in extreme cases. During a year where the Californian region has faced wildfires due to the impacts of climate change, it is pivotal to understand the precipitation patterns as it has major implications towards the surface environment in the region. Soil Moisture and Snow Water Equivalent (SWE) have been found to correlate with AR in the past and this study will further explore this relationship for the 2009-2010 time period. This study aims to predict precipitation with the help of AR indices, Soil Moisture and SWE in the state of California. In the preliminary results, random forest and support vector machine models have been fitted with crossvalidation and are able to predict if precipitation can be expected on a particular day with approximately 70-80% accuracy 6 months in advance. More years of analysis and a deeper analysis about the amount of precipitation expected should be further explored to understand the link between the different phenomena and factors affecting predictability.