Case Study: NASA achieves near-real-time detection of forest disturbances with MathWorks

NASA Develops Early Warning System for Detecting Forest Disturbances

NASA built ForWarn to provide a near–real-time early warning system that detects forest disturbances—caused by insects, drought, storms, blights, and wildfires—by analyzing multispectral MODIS satellite imagery. To meet the challenge of processing terabytes of multidimensional time-series data and rapidly prototyping algorithms, NASA used MathWorks tools, primarily MATLAB and its Image Processing, Mapping, Signal Processing, Optimization toolboxes, and MATLAB Compiler.

Using MathWorks’ MATLAB, NASA implemented the Time Series Product Tool (TSPT) and the Phenological Parameters Estimation Tool (PPET) to generate NDVI and other indices, remove clouds and outliers, perform curve fitting, and produce standalone executables. The MathWorks-based solution processes MODIS data every eight days, enabled detection of previously unnoticed hail damage, cut development time from years to hours for prototyping, and delivered reusable production software now used by a growing community of over 7,000 users.

NASA

Duane Armstrong

Chief of the Advanced Technology & Technology Transfer branch