QUANTIFYING GRASSLAND NON-PHOTOSYNTHETIC VEGETATION BIOMASS USING REMOTE SENSING DATA

Abstract

Non-photosynthetic vegetation (NPV) refers to vegetation that cannot perform a photosynthetic function. NPV, including standing dead vegetation and surface plant litter, plays a vital role in maintaining ecosystem function through controlling carbon, water and nutrient uptake as well as natural fire frequency and intensity in diverse ecosystems such as forest, savannah, wetland, cropland, and grassland. Due to its ecological importance, NPV has been selected as an indicator of grassland ecosystem health by the Alberta Public Lands Administration in Canada. The ecological importance of NPV has driven considerable research on quantifying NPV biomass with remote sensing approaches in various ecosystems. Although remote images, especially hyperspectral images, have demonstrated potential for use in NPV estimation, there has not been a way to quantify NPV biomass in semiarid grasslands where NPV biomass is affected by green vegetation (PV), bare soil and biological soil crust (BSC). The purpose of this research is to find a solution to quantitatively estimate NPV biomass with remote sensing approaches in semiarid mixed grasslands. Research was conducted in Grasslands National Park (GNP), a parcel of semiarid mixed prairie grassland in southern Saskatchewan, Canada. Multispectral images, including newly operational Landsat 8 Operational Land Imager (OLI) and Sentinel-2A Multi-spectral Instrument (MSIs) images and fine Quad-pol Radarsat-2 images were used for estimating NPV biomass in early, middle, and peak growing seasons via a simple linear regression approach. The results indicate that multispectral Landsat 8 OLI and Sentinel-2A MSIs have potential to quantify NPV biomass in peak and early senescence growing seasons. Radarsat-2 can also provide a solution for NPV biomass estimation. However, the performance of Radarsat-2 images is greatly affected by incidence angle of the image acquisition. This research filled a critical gap in applying remote sensing approaches to quantify NPV biomass in grassland ecosystems. NPV biomass estimates and approaches for estimating NPV biomass will contribute to grassland ecosystem health assessment (EHA) and natural resource (i.e. land, soil, water, plant, and animal) management.