You are here: Home Institutions SA Carbon Sinks Atlas SA Carbon Sinks Atlas Metadata

Metadata

Soil Organic Carbon (SOC) represents all the organic carbon in the soil to a depth of 1m. SOC is derived from the data provided by the Africa Soil Information System (AfSIS). In this case the organic carbon of the top 300mm of the natural soil was NOT reduced by land cover factors. Units: average gC/m2within 1km x 1km pixel
Soil Organic Carbon (SOC) represents all the organic carbon in the soil to a depth of 1m. SOC is derived from the data provided by the Africa Soil Information System (AfSIS). The organic carbon of soil is reduced by various land uses and agricultural practices and to account for this, the carbon in the top 300mm of the natural soil (SOC0-300) derived from AfSIS was reduced by a factor of 0.2 – 0.5 using the fractional land cover data per 1 X 1 km pixel. Units: average gC/m2within 1km x 1km pixel
Above Ground Herbaceous Biomass (AGBherb) is predominantly grasses, but also forbs, restios, sedges. It is based on published relationships between rainfall and yearly grass production, reduced proportionately to take into account competition by trees (TCF). AGBherbvaries greatly through the year – reaching a peak near the end of the growing and declining to near zero by the beginning of spring, especially in the presence of fire and/or herbivory. Units: average gC/m2within 1km x 1km pixel
The Total Biomass Organic Carbon of Natural and Transformed areas could simply be added together as they have each been corrected for fractional land cover. It therefore represents all the biomass organic carbon of natural vegetation, crops, urban areas and forestry plantations. For details see Total Biomass Organic Carbon of Transformed Areas and Total Biomass Organic Carbon of Natural Areas
Total Biomass Organic Carbon (TBOC) of natural land cover is the total sum of Above Ground Woody Biomass (AGBwoody), Below Ground Woody Biomass (BGBwoody), Above Ground Herbaceous Biomass (AGBherb), Below Ground Herbaceous Biomass (BGBherb), and Aboveground litter (AGL). In other words it is TEOC minus Soil Organic Carbon (SOC). Units: average gC/m2within 1km x 1km pixel
Total Ecosystem Organic Carbon (TEOC) is the total sum of Soil Organic Carbon (SOC) and Total Biomass Organic Carbon (of natural and transformed areas – crops and plantations) which includes Above Ground Woody Biomass (AGBwoody), Below Ground Woody Biomass (BGBwoody), Above Ground Herbaceous Biomass (AGBherb), Below Ground Herbaceous Biomass (BGBherb), and Aboveground litter (AGL). Carbon stocks were calculated to represent the long-term mean conditions. Units: average gC/m2within 1km x 1km pixel
Gross primary production (GPP) is the amount of biomass (units: gC/m2/yr) that plants create in a one year by capturing sunlight and CO₂ during photosynthesis. A production efficiency model was used to estimate GPP across South Africa. Carbon fluxes were calculated to represent the long-term mean conditions. The gross and net primary production (GPP, NPP) was calculated monthly and summed to a year using climatology of monthly weather for the period 1960 to 1990, and mean monthly FAPAR and PAR for period 2000 to 2012. There are thus 12 input files for each term, corresponding to the twelve months. The climatology averaging periods for climate and satellite data are different. For Fraction of Absorbed Photosynthetically Active Radiation (FAPAR), the MEdium Resolution Imaging Spectrometer (MERIS sensor) dataset covers the period 2000 to 2012, and the Photosynthetically Active Radiation (PAR) dataset is also for this period. Units: gC/m2/yr
Net primary production (NPP) is GPP minus autotrophic respiration (Ra). NPP = GPP – Ra In the version of NPP and Ra was assumed to be 0.5 GPP. A more elaborate calculation of Ra is under development by CSIR.
Above Ground Herbaceous Biomass (AGBherb) (MODIS version) is predominantly grasses, but also forbs, restios, sedges. It is based on the relationship between biomass (dry matter) and the leaf area index (LAI) using field data collected in the Lowveld region. The power empirical model based on LAI explained about 62% of variation and the model is as follows; ABGherb (gC/sq.m) = 264.37 x LAI^0.6374 The first step was to determine the maximum or peak productivity, as determined by the maximum biomass values in the month of January and February, based on the 13 year MODIS time series biomass data generated by CSIR. Finally, the ABGherb was determined by deriving the half of the averaged computed biomass, and to determine herbaceous carbon content (gC/sq.m, resulting biomass values were multiplied by 0.5. Units: gram of biomass (herbaceous) per square meter at 500m x 500m resolution
Woody fraction cover: The woody cover fraction is the fraction of area (0 to 1) projected on a horizontal plane occupied by woody plants. The product was developed through the integration of 2010 ALOS PALSAR-1 synthetic aperture radar images and LiDAR tracks. The LiDAR tracks were processed to derive a canopy height model for woody vegetation above 1 m at 1m pixel size, and to generate a detailed LiDAR woody cover product at 25m pixel size. The dual-polarized (HV, HH) SAR bands were modelled using the LiDAR woody cover as reference data for calibration and validation of the final SAR woody fraction cover map. Units: Percent Woody fraction cover
Above ground woody biomass: The above ground woody biomass is the total dry biomass of woody plants above 1m height and is expressed in tonnes per hectare. The product was developed through the integration of 2010 ALOS PALSAR-1 synthetic aperture radar images, LiDAR tracks, and field data of woody biomass. The LiDAR tracks were processed to derive a canopy height model for woody vegetation above 1 m at 1m pixel size. A detailed LiDAR aboveground woody biomass product was generated at 25m pixel size using LiDAR woody cover and height products and field data. The dual-polarized (HV, HH) SAR bands were modelled using the LiDAR woody aboveground biomass as reference data for calibration and validation of the final SAR woody aboveground biomass map. Units: Above ground (woody) biomass (tonnes) per hectare
The 2013-14 South African National Land-cover dataset produced by GEOTERRAIMAGE as a commercial data product has been generated from digital, multi-seasonal Landsat 8 multispectral imagery, acquired between April 2013 and March 2014. In excess of 600 Landsat images were used to generate the land-cover information, based on an average of 8 different seasonal image acquisition dates, within each of the 76 x image frames required to cover South Africa. The land-cover dataset, which covers the whole of South Africa, is presented in a map-corrected, raster format, based on 30x30m cells equivalent to the image resolution of the source Landsat 8 multi-spectral imagery. The dataset contains 72 x land-cover / use information classes, covering a wide range of natural and man-made landscape characteristics. The original land-cover dataset was processed in UTM (north) / WGS84 map projection format based on the Landsat 8 standard map projection format as provided by the USGS. For the purposes of the Carbon Atlas, the original 72 land cover classes were reclassified to 12 broad parent classes according to the definitions in Appendix A and hierarchical classes in Appendix E of GTI's Data User and Metadata Report. The colour scheme is similar to the colours used in the 72 original classes. This new layer serves as a reference for the Carbon layers and was resampled to an equivalent resolution of 1 km. Re-projection from UTM35 to Albers Conic Equal Area projection WGS84 was done.
Biomes coverage that groups the vegetation types of South Africa, Lesotho and Swaziland at the highest level. This represents a simplification of the biome level groupings of vegetation types
In the context of the Land Degradation Assessment in Drylands (LADA) project, the land use system approach for land degradation assessment has as guiding principle that land use is the major driving force of land degradation. Mapping of land use systems has therefore becomes a major activity within the LADA project at global and national level where land use units are considered the basic units in which land degradation and land improvements are mapped. The following spatial layers were found to be reliable enough and at an acceptable level of detail to be used in the production of the LUS map for South Africa: • National Land Cover 2000 (NLC Consortium); • Protected Areas in South Africa 2001 (Biodiversity GIS, South African Biodiversity Institute- SANBI); • VEGMAP 2006 (Mucina & Rutherford, SANBI); • Local Municipalities 2005 (Demarcation Board of South Africa). The LADA Land Use Map for South Africa has 18 land use/cover classes as a result of combining the 4 spatial layers discussed above. In combination with administrative boundaries a total number of 2447 unique combinations (which were equal to the number of LADA-WOCAT Mapping Questionnaires (QM) to be completed) were derived by integrating the Local Municipality (LM) boundaries with the Land Use Systems data.
Land degradation is a complex set of processes of impoverishment of terrestrial ecosystems mainly under the impact of human activities. Land degradation can be understood as the gradual or permanent loss of productivity of the land to produce ecosystem goods and services. During the Land Degradation Assessment in Drylands (LADA) project, the South African National Assessment of Land Degradation and Conservation was done between 2008 and 2011. 728 Contributing specialists throughout the country contribute their knowledge and experience on land degradation and sustainable land management during a series of 34 Participatory Expert Assessment Workshops. It is important to notice that the results of these national assessments are qualitative in nature and based on the perceptions of contributing specialists based on the assumption that they all know their respective assessment areas well enough to report on land degradation and conservation attributes. Variables on the state of land degradation (the extend, degree and rate of degradation processes) were combined with the level of impact of these degradation processes on ecosystem services to provide an unique Degradation Index (DI) for each type of degradation identified by the contributing specialist for each mapping unit as defined by the LADA Land Use Map for South Africa. The DI values range from 0 to 100, for all degradation types identified. The higher the DI value, the more degraded the area are.
Conservation activities are usually a response to land degradation problems and associated with the implementation of a single or combination of conservation measures. The purpose of conservation measures are either to prevent degradation before it occur, as a mitigation measure when full rehabilitation is still possible or for rehabilitation purposes in cases of serious land degradation. During the Land Degradation Assessment in Drylands (LADA) project, the South African National Assessment of Land Degradation and Conservation was done between 2008 and 2011. 728 Contributing specialists throughout the country contribute their knowledge and experience on land degradation and sustainable land management during a series of 34 Participatory Expert Assessment Workshops. It is important to notice that the results of these national assessments are qualitative in nature and based on the perceptions of contributing specialists based on the assumption that they all know their respective assessment areas well enough to report on land degradation and conservation attributes. During the Participatory Expert Assessment Workshops, contributing specialists identified for each mapping unit the main conservation groups and measures currently implemented by land users to address or respond to land degradation problems. As for the degradation assessment, variables describing these conservation measures were also combined to develop a unique Conservation Index (CI). Variables used were the extend of conservation measures implemented, the level of effective implementation of these measures, the effectiveness trend of these measures over time (last 10 years) and the level of impact of these conservation measures on ecosystem services. The CI values range from 0 to 100, for each mapping unit as defined by the LADA Land Use Map for South Africa. The higher the CI value, the more conservation measures are in place, the more effective they are, and the better the situation.
During the Land Degradation Assessment in Drylands (LADA) project, the South African National Assessment of Land Degradation and Conservation was done between 2008 and 2011. 728 Contributing specialists throughout the country contribute their knowledge and experience on land degradation and sustainable land management during a series of 34 Participatory Expert Assessment Workshops. It is important to notice that the results of these national assessments are qualitative in nature and based on the perceptions of contributing specialists based on the assumption that they all know their respective assessment areas well enough to report on land degradation and conservation attributes. One of the land degradation types considered during the LADA National Assessment was the reduction in vegetation cover and the increase in bare and unprotected soil. Since a positive correlation exist between Soil Organic Carbon Stocks (SOCS) and vegetation cover, areas where loss of cover has been identified by contributing specialists as one of the main degradation types, could have a negative impact on the available SOCS.
The biomes of South Africa are simulated using a dynamic vegetation model – the aDGVM- designed specifically for tropical and subtropical African ecosystems. Biomes distribution is simulated using climate simulations. Simulations were forced with projected changes in climate given by the Max Planck Institute for Meteorology's (Hamburg) ECHAM5 IPCC projections with atmospheric CO2 from IPCC (2007) SRES A1B projections. Biomes distribution are output for 1990.
The biomes of South Africa are simulated using a dynamic vegetation model – the aDGVM- designed specifically for tropical and subtropical African ecosystems. Biomes distribution is simulated using climate simulations. Simulations were forced with projected changes in climate given by the Max Planck Institute for Meteorology's (Hamburg) ECHAM5 IPCC projections with atmospheric CO2 from IPCC (2007) SRES A1B projections. Biomes distribution are output for 2100.
Soil erosion is an important form of land degradation and is among South Africa's most critical environmental issues. Gully erosion (in addition to sheet and rill erosion) has been identified as being a major source of sediment entering waterways in many catchments in South Africa (SA). The main purpose of creating this data was to establish the spatial extent of gully erosion per Province in South Africa. The aim was achieved by accurately mapping gully erosion features using SPOT 5 satellite imagery at a scale of 1:10 000. This data was created by means of visual interpretation and vectorization of SPOT 5 imagery (panchromatic sharpened images (2.5 m resolution) merged with multispectral bands (10 m resolution) to produce a 5 m resolution images) from various acquisition dates in 2006 to 2008. The outer boundary of gullies were delineated from SPOT 5 satellite images in ArcMap. Gullies were identified by their tone, shape, drainage pattern and association. The data provide useful information for conservation of natural resources projects of the Department of Agriculture, Forestry and Fisheries.
Above Ground Woody Biomass (AGB woody) is the total dry biomass of woody plants above 0.5m height and is expressed in tonnes per hectare. The product was developed through the integration of 2010 ALOS PALSAR-1 synthetic aperture radar images, LiDAR tracks, ancillary digital elevation model (DEM) derived variables and field data of woody biomass. Commercial plantations of very high AGB, permanent water bodies and urban areas were masked out since it was not captured by the LiDAR tracks. The mapped product was clipped to the savannah biome (Mucina and Rutherford 2006) and the SAR AGB woody map was resampled to 1km pixel resolutions using a mean aggregation method.
Above Ground Woody Biomass (AGB woody) is the total dry biomass of woody plants above 0.5m height and is expressed in tonnes per hectare. The product was developed through the integration of 2010 ALOS PALSAR-1 synthetic aperture radar images, LiDAR tracks, ancillary digital elevation model (DEM) derived variables and field data of woody biomass. Commercial plantations of very high AGB were masked out since it was not captured by the LiDAR tracks. The mapped product was clipped to the savannah biome (Mucina and Rutherford 2006) and the SAR AGBwoodymap was resampled to 100m pixel resolutions using a mean aggregation method.
Woody fractional cover is defined as the percentage area covered by a canopy surface over a particular pixel resolution and is expressed in percent (%). The product was developed through the integration of 2010 ALOS PALSAR-1 Synthetic Aperture Radar images (SAR), ancillary digital elevation model (DEM) derived variables and airborne LiDAR data. Commercial plantations of very high AGB, permanent water bodies and urban areas were masked out since it was not captured by the LiDAR tracks. The mapped product was clipped to the savannah biome (Mucina and Rutherford 2006) and the SAR AGB woody map was resampled to 1km pixel resolutions using a mean aggregation method.
Woody fractional cover is defined as the percentage area covered by a canopy surface over a particular pixel resolution and is expressed in percent (%). The product was developed through the integration of 2010 ALOS PALSAR-1 Synthetic Aperture Radar images (SAR), ancillary digital elevation model (DEM) derived variables and airborne LiDAR data. Commercial plantations of very high AGB, permanent water bodies and urban areas were masked out since it was not captured by the LiDAR tracks. The mapped product was clipped to the savannah biome (Mucina and Rutherford 2006) and the SAR AGB woody map was resampled to 100m pixel resolutions using a mean aggregation method.
The 2013-14 South African National Land-cover dataset produced by GEOTERRAIMAGE as a commercial data product has been generated from digital, multi-seasonal Landsat 8 multispectral imagery, acquired between April 2013 and March 2014. In excess of 600 Landsat images were used to generate the land-cover information, based on an average of 8 different seasonal image acquisition dates, within each of the 76 x image frames required to cover South Africa. The land-cover dataset, which covers the whole of South Africa, is presented in a map-corrected, raster format, based on 30x30m cells equivalent to the image resolution of the source Landsat 8 multi-spectral imagery. The dataset contains 72 x land-cover / use information classes, covering a wide range of natural and man-made landscape characteristics. The original land-cover dataset was processed in UTM (north) / WGS84 map projection format based on the Landsat 8 standard map projection format as provided by the USGS.
The modelled current long term annual climate conditions for the 30-year period 1986-2015. The total daily rainfall values simulated by six different global circulation models (GCMs), i.e. csiro, gfdl20, gfdl21, miroc, mpi, and ukmo were downscaled with the Cubic Conformal Atmospheric Model (CCAM) developed by CSIRO (Australia) and updated and modified by the CSIR (South Africa). The daily rainfall values were summed to yearly totals and the mean annual total rainfall were calculated for each downscaled global circulation model for the 30 year period. The median values are the 50th percentile values of the mean annual rainfall values for the six downscaled GCMs. Data are in raster format with a 0.5º resolution using a projected latitude / longitude coordinate reference system with `+proj=utm +zone=48 +datum=WGS84 +ellps=WGS84 +towgs84=0,0,0`.
The current long term annual climate conditions for the 30-year period 1986-2015. The total daily rainfall values simulated by six different global circulation models (GCMs), i.e. csiro, gfdl20, gfdl21, miroc, mpi, and ukmo were downscaled with the Cubic Conformal Atmospheric Model (CCAM) developed by CSIRO (Australia) and updated and modified by the CSIR (South Africa). The daily rainfall values were summed to yearly totals and the mean annual total rainfall were calculated for each downscaled global circulation model for the 30 year period. The 10th percentile values are based on the mean annual rainfall values for the six downscaled GCMs. Data are in raster format with a 0.5º resolution using a projected latitude / longitude coordinate reference system with `+proj=utm +zone=48 +datum=WGS84 +ellps=WGS84 +towgs84=0,0,0`.
The current long term annual climate conditions for the 30-year period 1986-2015. The total daily rainfall values simulated by six different global circulation models (GCMs), i.e. csiro, gfdl20, gfdl21, miroc, mpi, and ukmo were downscaled with the Cubic Conformal Atmospheric Model (CCAM) developed by CSIRO (Australia) and updated and modified by the CSIR (South Africa). The daily rainfall values were summed to yearly totals and the mean annual total rainfall were calculated for each downscaled global circulation model for the 30 year period. The 90th percentile values are based on the mean annual rainfall values for the six downscaled GCMs. Data are in raster format with a 0.5º resolution using a projected latitude / longitude coordinate reference system with `+proj=utm +zone=48 +datum=WGS84 +ellps=WGS84 +towgs84=0,0,0`.
The current long term annual climate conditions for the 30-year period 1986-2015. The maximum daily temperature values simulated by six different global circulation models (GCMs), i.e. csiro, gfdl20, gfdl21, miroc, mpi, and ukmo were downscaled with the Cubic Conformal Atmospheric Model (CCAM) developed by CSIRO (Australia) and updated and modified by the CSIR (South Africa). The mean maximum daily temperature was calculated per year for each of the six downscaled GCMs for the 30-year period. The median (50th percentile) values are based on the mean 30-year maximum daily temperature per year from the the six downscaled GCMs. Data are in raster format with a 0.5º resolution using a projected latitude / longitude coordinate reference system with `+proj=utm +zone=48 +datum=WGS84 +ellps=WGS84 +towgs84=0,0,0`.
The current long term annual climate conditions for the 30-year period 1986-2015. The maximum daily temperature values simulated by six different global circulation models (GCMs), i.e. csiro, gfdl20, gfdl21, miroc, mpi, and ukmo were downscaled with the Cubic Conformal Atmospheric Model (CCAM) developed by CSIRO (Australia) and updated and modified by the CSIR (South Africa). The mean maximum daily temperature was calculated per year for each of the six downscaled GCMs for the 30-year period. The 10th percentile values are based on the mean 30-year maximum daily temperature per year from the six downscaled GCMs. Data are in raster format with a 0.5º resolution using a projected latitude / longitude coordinate reference system with `+proj=utm +zone=48 +datum=WGS84 +ellps=WGS84 +towgs84=0,0,0`.
The current long term annual climate conditions for the 30-year period 1986-2015. The maximum daily temperature values simulated by six different global circulation models (GCMs), i.e. csiro, gfdl20, gfdl21, miroc, mpi, and ukmo were downscaled with the Cubic Conformal Atmospheric Model (CCAM) developed by CSIRO (Australia) and updated and modified by the CSIR (South Africa). The mean maximum daily temperature was calculated per year for each of the six downscaled GCMs for the 30-year period. The 90th percentile values are based on the mean 30-year maximum daily temperature per year from the six downscaled GCMs. Data are in raster format with a 0.5º resolution using a projected latitude / longitude coordinate reference system with `+proj=utm +zone=48 +datum=WGS84 +ellps=WGS84 +towgs84=0,0,0`.
The current long term annual climate conditions for the 30-year period 1986-2015. The minimum daily temperature values simulated by six different global circulation models (GCMs), i.e. csiro, gfdl20, gfdl21, miroc, mpi, and ukmo were downscaled with the Cubic Conformal Atmospheric Model (CCAM) developed by CSIRO (Australia) and updated and modified by the CSIR (South Africa). The mean minimum daily temperature was calculated per year for each of the six downscaled GCMs for the 30-year period. The median (50th percentile) values are based on the mean 30-year minimum daily temperature per year from the the six downscaled GCMs. Data are in raster format with a 0.5º resolution using a projected latitude / longitude coordinate reference system with `+proj=utm +zone=48 +datum=WGS84 +ellps=WGS84 +towgs84=0,0,0`.
The current long term annual climate conditions for the 30-year period 1986-2015. The minimum daily temperature values simulated by six different global circulation models (GCMs), i.e. csiro, gfdl20, gfdl21, miroc, mpi, and ukmo were downscaled with the Cubic Conformal Atmospheric Model (CCAM) developed by CSIRO (Australia) and updated and modified by the CSIR (South Africa). The mean minimum daily temperature was calculated per year for each of the six downscaled GCMs for the 30-year period. The 10th percentile values are based on the mean 30-year minimum daily temperature per year from the six downscaled GCMs. Data are in raster format with a 0.5º resolution using a projected latitude / longitude coordinate reference system with `+proj=utm +zone=48 +datum=WGS84 +ellps=WGS84 +towgs84=0,0,0`.
The current long term annual climate conditions for the 30-year period 1986-2015. The minimum daily temperature values simulated by six different global circulation models (GCMs), i.e. csiro, gfdl20, gfdl21, miroc, mpi, and ukmo were downscaled with the Cubic Conformal Atmospheric Model (CCAM) developed by CSIRO (Australia) and updated and modified by the CSIR (South Africa). The mean minimum daily temperature was calculated per year for each of the six downscaled GCMs for the 30-year period. The 90th percentile values are based on the mean 30-year minimum daily temperature per year from the six downscaled GCMs. Data are in raster format with a 0.5º resolution using a projected latitude / longitude coordinate reference system with `+proj=utm +zone=48 +datum=WGS84 +ellps=WGS84 +towgs84=0,0,0`.
% change in the long term mean annual total rainfall between 1986-2015 and 2036 - 2065. The mean annual rainfall for each downscaled global circulation model (GCM), i.e. csiro, gfdla20, gfdl21, miroc, mpi and ukmo using the the Cubic Conformal Atmospheric Model (CCAM) developed by CSIRO (Australia) and updated and modified by the CSIR (South Africa) was calculated for each of the 30 year periods. The percentage difference in the 30-year mean annual rainfall between the two periods was calculated for each of the six models as follows: ((rain2050 - rain2000)/rain2000)x100. The 50th percentile (median value) is based on these differences. Data are in raster format with a 0.5º resolution using a projected latitude / longitude coordinate reference system with `+proj=utm +zone=48 +datum=WGS84 +ellps=WGS84 +towgs84=0,0,0`.
% change in the long term mean annual total rainfall between 1986-2015 and 2036 - 2065. The mean annual rainfall for each downscaled global circulation model (GCM), i.e. csiro, gfdla20, gfdl21, miroc, mpi and ukmo using the the Cubic Conformal Atmospheric Model (CCAM) developed by CSIRO (Australia) and updated and modified by the CSIR (South Africa) was calculated for each of the 30 year periods. The percentage difference in the 30-year mean annual rainfall between the two periods was calculated for each of the six models as follows: ((rain2050 - rain2000)/rain2000)x100. The 10th percentile is based on these differences. Data are in raster format with a 0.5º resolution using a projected latitude / longitude coordinate reference system with `+proj=utm +zone=48 +datum=WGS84 +ellps=WGS84 +towgs84=0,0,0`.
% change in the long term mean annual total rainfall between 1986-2015 and 2036 - 2065. The mean annual rainfall for each downscaled global circulation model (GCM), i.e. csiro, gfdla20, gfdl21, miroc, mpi and ukmo using the the Cubic Conformal Atmospheric Model (CCAM) developed by CSIRO (Australia) and updated and modified by the CSIR (South Africa) was calculated for each of the 30 year periods. The percentage difference in the 30-year mean annual rainfall between the two periods was calculated for each of the six models as follows: ((rain2050 - rain2000)/rain2000)x100. The 90th percentile is based on these differences. Data are in raster format with a 0.5º resolution using a projected latitude / longitude coordinate reference system with `+proj=utm +zone=48 +datum=WGS84 +ellps=WGS84 +towgs84=0,0,0`.
ºC change in maximum daily temperature per year between 1986-2015 and 2036 - 2065. The mean annual maximum daily temperature for each downscaled global circulation model (GCM), i.e. csiro, gfdla20, gfdl21, miroc, mpi and ukmo using the Cubic Conformal Atmospheric Model (CCAM) developed by CSIRO (Australia) and updated and modified by the CSIR (South Africa) was calculated for each of the 30 year periods. The absolute difference between the two periods is based on the mean annual maximum daily temperature value and is calculated as follows: (Tmax2050-Tmax2000) for each of the six downscaled GCM models. The 50th percentile (median value) is based on these differences. Data are in raster format with a 0.5º resolution using a projected latitude / longitude coordinate reference system with `+proj=utm +zone=48 +datum=WGS84 +ellps=WGS84 +towgs84=0,0,0`.`.
ºC change in maximum daily temperature per year between 1986-2015 and 2036 - 2065. The mean annual maximum daily temperature for each downscaled global circulation model (GCM), i.e. csiro, gfdla20, gfdl21, miroc, mpi and ukmo using the Cubic Conformal Atmospheric Model (CCAM) developed by CSIRO (Australia) and updated and modified by the CSIR (South Africa) was calculated for each of the 30 year periods. The absolute difference between the two periods is based on the mean annual maximum daily temperature value and is calculated as follows: (Tmax2050-Tmax2000) for each of the six downscaled GCM models. The 10th percentile is based on these differences. Data are in raster format with a 0.5º resolution using a projected latitude / longitude coordinate reference system with `+proj=utm +zone=48 +datum=WGS84 +ellps=WGS84 +towgs84=0,0,0`.
ºC change in maximum daily temperature per year between 1986-2015 and 2036 - 2065. The mean annual maximum daily temperature for each downscaled global circulation model (GCM), i.e. csiro, gfdla20, gfdl21, miroc, mpi and ukmo using the Cubic Conformal Atmospheric Model (CCAM) developed by CSIRO (Australia) and updated and modified by the CSIR (South Africa) was calculated for each of the 30 year periods. The absolute difference between the two periods is based on the mean annual maximum daily temperature value and is calculated as follows: (Tmax2050-Tmax2000) for each of the six downscaled GCM models. The 90th percentile is based on these differences. Data are in raster format with a 0.5º resolution using a projected latitude / longitude coordinate reference system with `+proj=utm +zone=48 +datum=WGS84 +ellps=WGS84 +towgs84=0,0,0`.
ºC change in minimum daily temperature per year between 1986-2015 and 2036 - 2065. The mean annual minimum daily temperature for each downscaled global circulation model (GCM), i.e. csiro, gfdla20, gfdl21, miroc, mpi and ukmo using the Cubic Conformal Atmospheric Model (CCAM) developed by CSIRO (Australia) and updated and modified by the CSIR (South Africa) was calculated for each of the 30 year periods. The absolute difference between the two periods is based on the mean annual minimum daily temperature value and is calculated as follows: (Tmin2050-Tmin2000) for each of the six downscaled GCM models. The 50th percentile (median value) is based on these differences. Data are in raster format with a 0.5º resolution using a projected latitude / longitude coordinate reference system with `+proj=utm +zone=48 +datum=WGS84 +ellps=WGS84 +towgs84=0,0,0`.
AGBwoodyis estimated using the product of tree cover fraction (TCF) and height (Hveg), which were both estimated with coarse resolution satellite data, as a proxy for the volume woody vegetation. This volume is converted to aboveground biomass using constant, biome-specific Biomass Calibration Factor (BCFbiome). AGBwoody= (Hveg* TCF) * BCFbiomeUnits: (Two AGBwoodyoutputs are provided in two units) 1. average gC/m2within 1km x 1km pixel 2. average tonne Dry Matter / hectare (tDM/ha) within 1km x 1km pixel
AGBwoodyis estimated using the product of tree cover fraction (TCF) and height (Hveg), which were both estimated with coarse resolution satellite data, as a proxy for the volume woody vegetation. This volume is converted to aboveground biomass using constant, biome-specific Biomass Calibration Factor (BCFbiome). AGBwoody= (Hveg* TCF) * BCFbiomeUnits: (Two AGBwoodyoutputs are provided in two units) 1. average gC/m2within 1km x 1km pixel 2. average tonne DryMatter/hectare (tDM/ha) within 1km x 1km pixel
ºC change in minimum daily temperature per year between 1986-2015 and 2036 - 2065. The mean annual minimum daily temperature for each downscaled global circulation model (GCM), i.e. csiro, gfdla20, gfdl21, miroc, mpi and ukmo using the Cubic Conformal Atmospheric Model (CCAM) developed by CSIRO (Australia) and updated and modified by the CSIR (South Africa) was calculated for each of the 30 year periods. The absolute difference between the two periods is based on the mean annual minimum daily temperature value and is calculated as follows: (Tmin2050-Tmin2000) for each of the six downscaled GCM models. The 10th percentile is based on these differences. Data are in raster format with a 0.5º resolution using a projected latitude / longitude coordinate reference system with `+proj=utm +zone=48 +datum=WGS84 +ellps=WGS84 +towgs84=0,0,0`.
ºC change in minimum daily temperature per year between 1986-2015 and 2036 - 2065. The mean annual minimum daily temperature for each downscaled global circulation model (GCM), i.e. csiro, gfdla20, gfdl21, miroc, mpi and ukmo using the Cubic Conformal Atmospheric Model (CCAM) developed by CSIRO (Australia) and updated and modified by the CSIR (South Africa) was calculated for each of the 30 year periods. The absolute difference between the two periods is based on the mean annual minimum daily temperature value and is calculated as follows: (Tmin2050-Tmin2000) for each of the six downscaled GCM models. The 90th percentile is based on these differences. Data are in raster format with a 0.5º resolution using a projected latitude / longitude coordinate reference system with `+proj=utm +zone=48 +datum=WGS84 +ellps=WGS84 +towgs84=0,0,0`.
In areas which have been transformed to cultivation or forestry plantations municipal level agricultural senses data were used to assign the biomass values to areas covered by the specific fractional land cover or crop type.
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
Document Actions