A System for Monitoring and Forecasting Land Surface Phenology Using Time Series of JPSS VIIRS Observations and Its Applications

Abstract:

Land surface phenology quantifies seasonal dynamics of vegetation properties including the timing and magnitude of vegetation greenness from satellite observations. Over the last decade, historical time series of AVHRR and MODIS data has been used to characterize the seasonal and interannual variation in terrestrial ecosystems and their responses to a changing and variable climate. The Visible Infrared Imaging Radiometer Suite (VIIRS) instrument on board the operational JPSS satellites provides land surface observations in a timely fashion, which has the capability to monitor phenological development in near real time. This capability is particularly important for assisting agriculture, natural resource management, and land modeling for weather prediction systems. Here we introduce a system to monitor in real time and forecast in the short term phenological development based on daily VIIRS observations available with a one-day latency. The system integrates a climatological land surface phenology from long-term MODIS data and available VIIRS observations to simulate a set of potential temporal trajectories of greenness development at a given time and pixel. The greenness trajectories, which are qualified using daily two-band Enhanced Vegetation Index (EVI2), are applied to identify spring green leaf development and autumn color foliage status in real time and to predict the occurrence of future phenological events. This system currently monitors vegetation development across the North America every three days and makes prediction to 10 days ahead. We further introduce the applications of near real time spring green leaf and fall color foliage. Specifically, this system is used for tracing the crop progress across the United States, guiding the field observations in US National Phenology Network, servicing tourists for the observation of color fall foliage, and parameterizing seasonal surface physical conditions for numerical weather prediction models.