Guerschman, JP; Hill, MJ; Leys, J; Heidenreich, S (2020). Vegetation cover dependence on accumulated antecedent precipitation in Australia: Relationships with photosynthetic and non-photosynthetic vegetation fractions. REMOTE SENSING OF ENVIRONMENT, 240, 111670.

The development of fractional vegetation cover products for Australia that resolve cover into photosynthetic vegetation (F-PV), non-photosynthetic vegetation (F-NPV) and bare soil/rock (F-BS) provides a new basis for examination of responses of vegetation cover to long term precipitation cycles, and to explore the interaction between these responses and land cover type, land use and other land surface properties. In this study, the relationship between accumulated antecedent precipitation (AAP) from 1 to 60 months and average monthly F-PV and F-NPV from the MODIS Fractional Cover Product is examined over a 17 year period from 2001 to 2018. The maximum R-2 value, regression coefficients for the maximum R-2, and number of accumulated months to the maximum R-2 were mapped for each month of the year for F-PV and F-NPV. Behaviour of responses in relation to land use, land cover, and soil water holding capacity was analysed based on pixel frequencies of classes at sub-region scale in the Interim Biogeographic Regionalisation of Australia. The analysis showed that F-PV is largely dependent upon AAP in the preceding 12 months, however responses to longer periods of AAP also occur in specific land use-vegetation type combinations. The study also showed that positive responses in F-NPV could be driven by AAP from as much as 60 months, but that F-NPV is reduced in many areas in response to increased AAP. The presence or absence of domesticated livestock grazing as defined by Australian land use mapping was a major influence on response to AAP of both F-PV and F-NPV with statistical analysis indicating interactions between major natural vegetation type and grazing. In highly responsive areas (R-2 > 0.6) monitoring of land condition could be enhanced by testing of regional cover levels for major deviations from the long terms responses that might suggest land management concerns.