Digitizing Historical Forest Service Data

When ecologists are working in the field, they often record their data on datasheets by hand. This hard-won information then tends to remain trapped in physical copies of datasheets which then get stored into filing cabinets, preventing further analysis. We are collaborating with the Sawtooth National Forest Service, which has collected decades of data on historical vegetation and soil conditions in the Sun Valley, Idaho area to digitize their historical data. The goal of this project is to create an Optical Character Recognition (OCR) model able to process the collected handwritten datasheets and generate a digitized version of them. By making nearly a century of environmental data ready for statistical analysis, this project will allow Forest Service and BSU scientists to answer important questions about how some of Idaho's most spectacular landscapes have been affected by climate change, sheep grazing, and natural resource management decisions across areas and timeframes that were previously impractical to tackle.