Analysing the effect of drought on net primary productivity of tropical rainforests in Queensland using MODIS satellite imagery

Abstract

The changing climate is characterised by increased frequency and severity of extreme weather events that include cyclones, drought, flooding, heatwaves and snowing. Such climatic patterns are predicted to have negative effects on tropical forests. Given the role played by these forests in hydrological cycle, carbon cycling and biodiversity, it is crucial to quantify these effects to ensure adequate monitoring and management of these forest ecosystems. The aim of this study was to analyse the effect of drought on the Net Primary Productivity (NPP) of tropical rainforests. The specific objectives were to a) determine existing trends of NPP over the study period, b) assess the relationship between NPP and climatic - biophysical factors, c) analyse the effect of drought on NPP, and d) evaluate inter-annual variation of NPP. Moderate Resolution Imaging Spectroradiometer (MODIS) satellite imagery, thematic data sets and climatic data sets were acquired and analysed using GIS techniques, image processing tools and statistical techniques. Correlation and regression analyses were used to establish relationship between NPP and climatic-biophysical factors. NPP had a positive trend over the study period even though there was a significant reduction of NPP in 2003 and 2006 as a result of drought events in 2002 and 2005, respectively. A significantly high r2 (0.8) was achieved when comparing rainfall of current year and NPP of subsequent year. In this study, it was concluded that rainfall and drought had a lag effect on NPP. The interannual variation of NPP in this region was significantly (P < 0.001) influenced by the rainfall amount (r = 0.8), maximum temperature (r = 0.8), solar radiation (r = 0.84) and potential evaporation (r = -0.61). The influence of temperature was evident within distinct periods (cycles). The index representing soil water holding capacity in the B horizon was found to have a positive relationship (r = 0.82) with NPP. The other biophysical factors (slope, DEM, aspect and terrain wetness index) did not exhibit strong correlations with NPP, indicating a weak influence. The influence exerted by the interaction of biophysical and climatic factors on NPP requires more research.