WAVES vegetation parameter sets for 21 South African vegetation types (WRC Report 1319/1/08)

This dataset was harvested from WRC Report No. 1319/1/08, which developed and tested simplified models of vegetation water use for a wide range of South African vegetation types. The project adapted the WAVES biophysical process model (CSIRO) into WAVES (SA), linking it to a South African climate database and adding flexible carbon allocation, spring translocation, frost and burning/harvest options. Field measurements of evapotranspiration (ET) were made or collated using scintillometry, Bowen ratio, eddy covariance and heat pulse sap flow techniques, and WAVES was calibrated against these observations at sites across the Grassland, Savanna, Thicket, Forest and Fynbos biomes, as well as wetlands, alien invasive plant stands, dryland crops and plantation forests. The core of the dataset is the full set of calibrated WAVES vegetation parameter values published in Appendix 1 of the report: 28 parameters (including canopy and soil albedo, rainfall interception and light extinction coefficients, maximum carbon assimilation rate, conductance slope, minimum soil water potential, growth temperature thresholds, maximum rooting depth, specific leaf area, respiration and leaf mortality rates, aerodynamic resistance, soil nutrient index and carbon allocation and frost switches) for 21 parameter sets representing 17 sites, with separate overstorey and understorey sets for savanna, valley thicket and indigenous forest. A site table was compiled for each parameter set, recording biome, district, dominant species, report section, coordinates and altitude where reported, mean annual precipitation, the WAVES simulation period and the weather stations used. Supporting tables from the report were also digitised, namely mean annual precipitation, runoff and estimated ET for 20 Drakensberg and KwaZulu-Natal grassland catchments, initial root carbon distributions with soil depth for grassland, savanna, forest and Fynbos, species composition and fortnightly leaf area and specific leaf area measurements for the Bellevue grassland, soil properties for Nylsvley and Noodsberg, plant carbon contents for Nylsvley, simulated site water balances for Noodsberg valley thicket and Groenkop forest, Pinus patula biomass calculations, and simulated versus Bowen ratio ET for an Acacia mearnsii plantation. The report was born-digital, so values were extracted from the embedded text layer and checked against page images. Arithmetic and plausibility checks identified a small number of internal inconsistencies in the source, which were retained unchanged and documented in a dedicated verification sheet. Data shown only in figures, such as daily ET time series, were not digitised. The parameter sets were intended as a starting point for simulating ET of similar vegetation elsewhere in South Africa and for land-use change scenario studies.