Land Surface Phenologies of Grasslands: Comparing VenμS Time Series from Naryn, Kyrgyzstan and the Eastern Sandhills of Nebraska, USA

For two grazed grassland sites with contrasting terrain, soils, and vegetation composition, we modeled the land surface phenology (LSP) as a convex quadratic (CxQ) function of accumulated growing degree-days calculated from MODIS 1 km land surface temperature data. We fitted time series of three complementary VIs: Gitelson’s chlorophyll index for erectophile canopies using the 742 nm red-edge band; the VENµS variant of the MERIS Terrestrial Chlorophyll Index; and the NDVI. Fitting was done separately for each pixel time series within the data frame, and it was done iteratively toward quality of fit targets. Four phenometrics from the CxQ models were used for comparisons: Peak Height (PH—the maximum model VI); Thermal Time to Peak (TTP—duration of the modeled green-up phase); Half-Time Value (HTV—VI at half TTP); and the early season AUC (Area Under the Curve above VI-specific bare soil threshold integrated from the first observation above the threshold to the peak height). Fits were primarily constrained by number of cloud-free observations but generally high. However, we detected a few anomalously high VI values associated with very low reflectance values, which occurred sporadically in space and time. Together the spectral bands, high spatial resolution, and quick tempo make VENµS a excellent sensor for characterization of land surface phenologies in grasslands.