Investigating the use of a leaf trait-based approach to remotely measure tropical forest productivity

Tropical forests and savannahs account for 60% of terrestrial carbon cycling. Thus, understanding rates of carbon sequestration in these areas, along with projected changes, is vital for predicting future climate. MODIS satellites offer two products to measure gross primary productivity (GPP) and net primary productivity (NPP) across vegetated areas. Yet significant doubt has been cast on the ability of these products to accurately estimate the magnitude and spatial patterns of GPP and NPP over tropical forests (Cleveland et al. 2015). It is clear that in these regions a new methodology is required. This proposal aims to investigate the use of a novel leaf trait-based approach to remotely sense tropical forest productivity. Using leaf spectroscopy, airborne hyperspectral data and partial least square regression analysis (PLSR), it will build on existing literature to show that carbon cycling processes, such as leaf photosynthesis, respiration and woody productivity, can be accurately estimated by calibrating SENTINEL satellite data on field-collected values. Ultimately, it intends to create maps of tropical forest productivity in Google Earth Engine (a cloud computing platform for processing satellite imagery) for public and scientific use. Initial research carried out for my MSc dissertation in this area was very promising, and, following a year-long position as Research Scholar at Northern Arizona University, where I collaborated with Prof. Chris Doughty to expand this work, I believe I am uniquely placed to undertake this research and advance this novel and emerging field.