Phenology analysis of moist decedous forest using time series Landsat-8 remote sensing data

Accurate and up–to-date monitoring of forests at mountainous regions at regular time interval is a challenging task. Remote sensing derived metrics such as vegetation indices are the most widely used tools for the estimation of forests phenological attributes and ecosystem monitoring. Thus, in this study, assessment of spectral traits has been carried out using satellite RS sensors derived information. In this study, the assessment was carried out from 2013 to the mid-2015 using Landsat-8 to understand NDVI phenology and modeling of phenology trends using temporal normalized phenology index (TNPI) and remote sensing derived variables such as land surface temperature (LST), elevation, aspect and slope within moist deciduous forest (MDF) of Doon valley of Western Himalayan of India. Results indicated that variations in topography and LST showed strong associations (p<0.001) with Landsat-8 derived TNPI. We observed that NDVI changed in lower elevation areas due to maximum change in LST. In conclusion, cross-validated statistics confirmed that TNPI was tested successfully for another MDF test site using Landsat-8 derived NDVI for two time steps of maximum and minimum vegetation growth period.