Welcome to the Albany CSTAR Data Fusion Winter Weather Project!

This page has been developed through research conducted as a part of a NOAA CSTAR grant

All plots and data displayed is experimental in nature and should not be used other than in a research framework

To learn more about the algorithm, please visit the Data and Methods section
For information about how to use this product, please visit the Training tab

NOAA LOGO
ALBANY LOGO
NYS Mesonet LOGO




About this project:

This project is a part of the University at Albany’s long-standing collaboration with the National Weather Service (NWS) through the Collaborative Science, Technology, and Applied Research (CSTAR) Program; this specific project is supported by NOAA Award Number NA19NWS4680006. To learn more about this and other CSTAR projects from the University at Albany, please visit the VLAB page.

The goal of the Data Fusion project is to merge a multitude of data sources forecasters can use to make a forecast and succinctly and effectively combine them to improve our forecasting ability of hazardous weather events like winter mixed precipitation events. To translate this work to operations, this website was created to share the live updating plots with winter precipitation probabilities. In addition to the website, training modules or quick references for the forecasters will be created as another tool to share what information and data is most important to determining precipitation type.

This tool uses machine learning as its base because of its ability to synthesize and learn how to differentiate between precipitation types. A random forest machine learning algorithm will be used in this case because of its ability to detail, clearly and explicitly, the decision-making process behind the algorithm, while handling large amounts of data. This tool is meant to be used in conjunction with a forecaster because the computer does not take things like local terrain into account, so local NWS forecasters can use this tool and their local knowledge to make the best possible prediction. They can also use the output of the most important variables to help when making their own assessment of what the precipitation type will be based on information in other locations.


About the Data Fusion Team:

Brian Filipiak is currently a graduate student at the University at Albany. His work on this project is supported by his University at Albany advisors: Kristen Corbosiero, Andrea Lang, Nick Bassill, and Ross Lazear.

In addition to the Data Fusion team at the University at Albany, Brian’s NWS focal points for the project are Christina Speciale, Neil Stuart, and Mike Evans at the Albany WFO. They have and continue to provide valuable guidance, suggestions, and feedback throughout the process.