Skip to main content
Dryad

Global dryland vegetation memory

Cite this dataset

Kusch, Erik; Seddon, Alistair W.; Davy, Richard (2022). Global dryland vegetation memory [Dataset]. Dryad. https://doi.org/10.5061/dryad.k98sf7m6d

Abstract

1. Vegetation memory describes the effect of antecedent environmental and ecological conditions on the present ecosystem state and has been proposed as an important proxy for vegetation resilience. In particular, strong vegetation-memory effects have been identified in dryland regions, but the factors underlying the spatial patterns of vegetation memory remain unknown.

2. We aim to map the components and drivers of vegetation memory in dryland regions using state-of-the-art climate-reanalysis data and refined approaches to identify vegetation-memory characteristics across dryland regions worldwide.

3. Using a framework which distinguishes between intrinsic and extrinsic ecological memory, we show that: (i) intrinsic memory is a much stronger component than extrinsic memory in the majority of dryland regions; and  (ii) climate reanalysis data sets change the detection of extrinsic vegetation memory effects in some global dryland regions.

4. Synthesis. Our study offers a global picture of the vegetation response to two climate forcing variables using satellite data, information which is potentially relevant for mapping components and properties of vegetation responses worldwide. However, the large differences in the spatial patterns in intrinsic vegetation memory in our study compared to previous analyses show the overall sensitivity of this component in particular to the initial choice of extrinsic forcing variables. As a result, we caution against using the oversimplified link between intrinsic vegetation-memory and vegetation recovery rates at large spatial scales.

Methods

NDVI data was downloaded from https://climatedataguide.ucar.edu/climate-data/ndvi-normalized-difference-vegetation-index-3rd-generation-nasagfsc-gimms using the gimms R package (https://cran.r-project.org/web/packages/gimms/index.html). The data was subsequently aggregated to monthly maximum raster layers and masked to dryland regions.

ERA5 data was downloaded from the Climate Data Store and downscaled using R.

Funding

Bjerknes Fast Track Initiative