ESTIMATING CANOPY NITROGEN AND FIBRE CONCENTRATION ACROSS C3 AND C4 GRASSLANDS USING WORLDVIEW-2 MULTISPECTRAL DATA AND THE RANDOM FOREST ALGORITHM
Clement Adjorlolo, Onisimo Mutanga and Moses Cho
KwaZulu-Natal Department of Agriculture and Environmental Affairs, Natural Resources, University of KwaZulu-Natal, School of Agricultural, Earth and Environmental Sciences, Council for Scientific and Industrial Research (CSIR), Natural Resources and the Environment

 

Abstract
Global climate change is expected to cause changes in the composition of plant functional types or, shifts in the percentage cover and abundance of grass species following the C3 and C4 photosynthetic pathways. These two groups differ in a number of physiological, structural and biochemical aspects. The C3 grass type generally contains a higher concentration of nitrogen (N) but lower carbon (C)/N ratio, compared to C4 grasses. Remote sensing applications have been proven to be a source of proximal data for estimating several canopy parameters related to biophysical, physiological or biochemical characteristics of vegetation. This study assesses the potential of WorldView-2 (WV2) satellite imagery to estimate and map the spatial distributions of nitrogen (N) and crude fibre concentration in montane grasslands of the Cathedral Peak region of the Drakensberg Mountain range, South Africa. Random forest (RF) regression algorithm was used to develop a relationship between two-band vegetation indices computed from the WV2 image and N, and fibre concentrations (n = 150, for each dataset). Using the validation datasets (n = 64), the RF regression explained 71% and 66% of the variation in predicting N and fibre content, respectively; across the C3 grass, Festuca costata, and C4 grasses, Themeda triandra and Rendlia altera grasslands. The RF modeling analyses yielded Root Mean Square Error of prediction for N = 0.24% compared to observed mean of 1.09% and fibre = 9.55% compared to observed mean of 64.16%. Overall, the results obtained suggest that new multispectral data with unique band setting, such as WV2, are capable of estimating grass N and fibre concentrations. This study demonstrated that canopy concentration of forage nutrient variables in a C3 and C4 grassland environment can be rapidly and non-destructively predicted and mapped using remote sensing data.

 
Presentation Topic
ESTIMATING CANOPY NITROGEN AND FIBRE CONCENTRATION ACROSS C3 AND C4 GRASSLANDS USING WORLDVIEW-2 MULTISPECTRAL DATA AND THE RANDOM FOREST ALGORITHM
Contact Mr Adjorlolo:
Email Clement Adjorlolo