The Land Surface Temperature (LST) product from the Visible Infrared Imager Radiometer Suite (VIIRS) will provide key information for estimating Earth surface energy and water fluxes, improving weather forecasting at high spatial and temporal resolutions, and monitoring climate change. The main challenges in estimating LST from VIIRS and similar sensors (e.g., AVHRR, MODIS, AATSR) are the need to correct for surface emissivity and atmospheric water vapor attenuation, and the difficulty in quantifying LST product accuracy due to the lack of comparable in situ measurements. This work presents a new methodology that combines in situ observations with fine-scale surface modeling to allow routine quantitative assessment of VIIRS LST products and a physically-based framework for testing and improving the retrieval algorithm. The validation scheme uses NOAA's surface observations (surface and meteorological) collected by the US Climate Reference Network (CRN) over a large range of climate regimes and land cover types in the US. The scaling methodology consists of the merging of information collected at different spatial resolutions and the SEtHyS land surface model to fully characterize the satellite products, i.e. measurements from ground stations to satellites platforms at high and moderate resolutions. The approach can be applied to ground stations worldwide. The project was initiated under NPOESS/JPSS program, and it is tested and proven here with NASA/MODIS data. The primary goal is to establish, in near real time, the accuracy of the LST product derived from VIIRS over a selection of field validation sites for the VIIRS algorithm working group and the science-user community.