Monitoring surface type changes with S-NPP/JPSS VIIRS observations

Accurate representation of currently actual land surface types at regional to global scales is critical for land surface parameterization in numerical weather, climate, hydrological and ecological models. Depending on the time scale, monitoring land surface type changes is also increasingly important for natural disaster assessment and natural resources management. To provide near real time surface type information for downstream data product generation (e.g. land surface temperature) from the Suomi-National Polar-orbiting Partnership (S-NPP) Visible and Infrared Radiometer Suite (VIIRS), numerical weather prediction models, and natural disaster assessment and resources management, the Surface Type (ST) Environmental Data Record (EDR) has been generated since early 2012 from the VIIRS observations. The ST EDR contains a static surface type label as well as current day active fire and snow conditions for each 750m pixel of the whole globe. In addition to the surface type changes caused by active fire and snow, other surface type changes may occur and have implications for land surface parameterization. For example, flooding of a land area will completely change its surface albedo, temperature, and the water, energy and CO2 exchange rates. In this study, a near real time surface type change monitoring framework is developed and tested to provide daily surface type information using the S-NPP and future Joint Polar Satellite System (JPSS) VIIRS surface reflectance observations. This near real time surface type information includes burned, flooded, deforested/urbanized, crop-harvested areas in addition to the surface type label described by a static surface type map generated from annual or multi-year satellite observations. In this paper, the current VIIRS surface type EDR and its characteristics are introduced as reference. The design of a surface type change monitoring framework and the surface type change detection methodology are described, followed by three case studies in which this framework is used to detect flooding and burned areas. The potential for making the framework operational for use in operations of numerical weather prediction, water resources and disaster management is also discussed.