MAPPING PLANT NITROGEN CONCENTRATION WITHIN LUVUVHU CATCHMENT AREA USING SENTINEL-2 IMAGERY

Unarine Emmanuel Tshishonga ^* Oupa Malahlela

• University of Venda, Thohoyandou, South Africa

Plant nitrogen (N) estimation using real-time and non-convection methods is very crucial for ecosystem management. This study aimed at estimating plant N concentration of indigenous vegetation in Luvuvhu River Catchment (LRC) using both field and remotely sensed data from Sentinel-2 imagery. The study used three different categories of spectral indices to fulfil its objectives. Red edge based, nitrogen related, and combined spectral indices based on Sentinel-2 data were subjected to stepwise regression on R studio software to determine which category of spectral indices is more efficient in estimating plant N concentration. Results have shown that combined spectral indices performed better with R 2 = 0.59, RMSE = 0.47% and MAE = 0.38%, followed by N based spectral indices with R 2 = 0.44, RMSE = 0.65% and MAE = 0.48%, and the last category is red edge based spectral indices with R 2 = 0.35, RMSE = 0.81% and MAE = 0.65%. The coefficients of the best performing model obtained from stepwise regression were used to compute multiple linear regression on QGIS to produce a map showing the concentration of plant N across the study area. Plant N varies with plant species and, the thematic map created show how plant N is distributed across the study area. With the help of this study, forest managers can better manage the natural vegetation by collaborating with forest communities. This possible partnership will create green jobs in addition to revenue. The link of the natural regeneration, reforestation, agroforestry and quantification by Sentinel-2 images for emission reduction will be beneficial for their livelihood. On a broader scale, participatory management is a good way to mitigate and adapt to climate change. On the other hand, the study suggests that more in-depth research should be conducted to explore further properties of red-edge indices for vegetation parameters prediction.