This data set was collected during February-March 2000 wet season and September 2000 dry season field campaigns of SAFARI 2000. Mongu in Zambia and Pandematenga (aka Kasane) and Tshane in Botswana were visited during the wet season campaign. Dry season data are for Mongu only. Hemispherical photographs, from which Plant Area Index (PAI) estimates are derived, were obtained at the field sites to characterize vegetation structural changes along the Kalahari Transect. At each site, photographs were collected using a Nikon Coolpix 950 digital camera with a fish-eye lens mounted on a tripod, every 25 m along three transects, each 750 m in length. The methodology of van Gardingen et al. (1999) was applied to process the data. The photographs are classified into sky and vegetation (trunk, green and senescent leaves, and branches) using an unsupervised classification scheme. Derivation of gap fraction estimates was carried out by dividing the azimuth sector for each view zenith into 72 equal area segments with the gap fraction estimated for each segment. The average azimuth segmented gap fraction was used to characterize the gap fraction at the current view zenith angle. This approach attempts to reduce the underestimation introduced by assuming a random distribution in clumped canopies. PAI estimates were derived from the gap fraction data using Miller's (1967) formula for average foliage density.Data are in comma-delimited ASCII files, one for each transect at each site, containing a 5-line header, then a blank line, column names, and the two-column data (sample number and PAI).

Data presented in this data set were collected during an intensive field campaign in Botswana between February 28 and March 18, 2000 along the Kalahari Transect as part of the SAFARI 2000 wet season field campaign. The sites visited were Pandamatenga, Maun, Okwa River Crossing, and Tshane (north to south). Individual leaf blade measurements were made on replicate samples from selected dominant and subdominant tree species using an optical lens and graticule. Leaves used in the study had recently-matured new growth, and were fully exposed to the sun for a significant part of the day. The data set is comprised of individual leaf blade dimensions along the length and width of each leaf by tree species as well as the mean of the replicate leaf length and width samples. The data are in comma-delimited ASCII format (kt_leaf_dimensions.csv).

This data set contains leaf area index (LAI), leaf inclination angle, and canopy dimension data from study sites along the Kalahari Transect in southwest Botswana. The data were collected during the 2001 wet season field campaign of the SAFARI 2000 at a total of seven plots of 200 x 150 meter dimensions: two plots each at Tshane and Mabuasehube and three plots at Tsabong. The data set consists of measurements of leaf angle for plot dominant woody species, LAI calculated from overstory and understory photosynthetically active radiation (PAR) measurements, and canopy dimension data (i.e., crown height, crown width, and height to crown) for grass and woody vegetation for use in the parameterization of plant canopy reflectance models. The data files are stored as ASCII table files, in comma-separated-value (.csv) format, with column headers. Photographs (.jpg) are provided of each plot to provide an idea of site conditions. The photographs can be viewed on the S2K Photo Gallery pages.

Data from the LAI-2000 instrument were processed to determine the leaf area index (LAI) at the EOS Validation Core Site in Kataba Local Forest, approximately 20 km south of Mongu, Zambia. Measurements began in 2000 and continued into 2002, with measurements collected about every month throughout the growing season to examine the phenology of LAI for the site. The LAI-2000 measures the intensity of blue light in five upward-looking concentric conical rings. Measurements are made under the forest canopy and compared with open-sky measurements to determine transmittance for each of the five viewing angles. The sensor head was placed at ground level while the sensors measured light levels in conical scans. Effective leaf area was calculated from the transmittance in the different view angles based on the assumption of a random distribution of leaves (Welles and Norman, 1991).The LAI-2000 was carried along three parallel transects, each 750 m long and spaced 250 m apart. Each transect was divided into 25 m segments, and measurements were collected at the endpoints of each segment. Data from all transects were combined to provide site-average LAI for each sampling date. The length and spacing of the transects were chosen to sample an area large enough to be representative of a 1 km MODIS pixel. Ground observations of LAI from this study compared with MODIS LAI products were found to be in close agreement.The data are stored in an ASCII text file, in comma-separated-value (csv) format, with column headers.

Data from the Tracing Radiation and Architecture of Canopies (TRAC) instrument were collected at five sites along the International Geosphere-Biosphere Programme (IGBP) Kalahari Transect, including Mongu in Zambia and Pandamatenga, Maun, Okwa River Crossing, and Tshane in Botswana, during the 2000 wet season field campaign (March-April) of SAFARI 2000. At the Mongu site, TRAC measurements began in August of 1999 and continued beyond the 2000 wet season field campaign, about every month for the rest of 2000. The TRAC instrument contains pyranometers that are sensitive to photosynthetically active radiation (PAR) at 400-700 nm. The TRAC measures the PAR flux transmitted through the overstory canopy continuously at 32 Hz. The parameters derived from the TRAC instrument include estimates of plant or leaf area index (PAI, LAI), overstory gap fraction, and clumping index.

The Boston University team collected several data sets along the Kalahari Transect during the SAFARI 2000 wet season field campaign between March 3 and March 18, 2000, to support the validation of the MODIS LAI/FPAR algorithm. Ground measurements of LAI, FPAR, leaf hemispherical reflectance and transmittance, and canopy transmittance were made using a LAI-2000 plant canopy analyzer, an AccuPAR ceptometer, a LiCor 1800-12S External Integrating Sphere (LI-1800) portable spectroradiometer, and an ASD handheld spectroradiometer. Leaf spectral data are provided in this data set. Leaf spectral measurements were made on samples from dominant tree, shrub, and grass species at 5 different Kalahari Transect sites - Mongu in Zambia and Pandamatenga, Maun, Okwa River, and Tshane in Botswana (from north to south)- where vegetation ranges from moist closed woodlands to arid sparsely-shrub-covered grasslands.

This data set provides species distribution, basal area, height, and crown cover of woody stems at 10 sites along the Kalahari Transect where a large gradient in both the mean and variation of annual rainfall results in dramatic changes in vegetation structure. Some of the data were collected during earlier Kalahari Transect projects in 1995 and 1997 at Vastrap, South Africa; Sandveld and Sachinga, Namibia; and Maziba, Senanga, and Lukulu, Zambia. The rest of the data were collected at Mongu, Zambia; and Pandamatenga, Maun, and Tshane, Botswana during the February-March 2000 wet season field campaign of SAFARI 2000. Stem maps were generated at each site using a variable-width belt-transect approach. Tree location, species, diameter, height, and major and minor axis of crown dimensions were measured for each individual taller than 1.5 meters. For multi-stemmed individuals, the diameter of each stem was recorded separately. Canopy area was calculated to be an ellipse defined by the two major axes of measurement. Canopy height was estimated using a clinometer. Biomass was calculated following Goodman (1990) as modified by Dowty (1999).There are two ASCII data files, in comma-delimited format. The stem map file contains records of living, dead, and cut stem allometry, canopy geometry, and biomass at the SAFARI sites. The species list file provides plant family, genus, and species names, numerical codes that correspond to the stem map file, and species common names in English and AFRICAANS.

The Boston University team collected several data sets along the Kalahari Transect during the SAFARI 2000 wet season field campaign between March 3 and March 18, 2000 to support the validation of the MODIS LAI/FPAR algorithm. Ground measurements of LAI, FPAR, leaf hemispherical reflectance and transmittance, and canopy transmittance were made using a LAI-2000 plant canopy analyzer, an AccuPAR ceptometer, a LiCor 1800-12S External Integrating Sphere (LI-1800) portable spectroradiometer, and an ASD handheld spectroradiometer. Leaf spectral data are provided in this data set. Leaf spectral measurements were made on samples from dominant tree, shrub, and grass species at 5 different Kalahari Transect sites - Mongu in Zambia and Pandamatenga, Maun, Okwa River, and Tshane in Botswana (from north to south) - where vegetation ranges from moist closed woodlands to arid sparsely-shrub-covered grasslands. Measurements were made on site with a LI-1800 portable spectroradiometer right after the leaves were cut from the trees or shrubs. Three or four sample leaves of each dominant species were measured.The data files, in ASCII comma-delimited (.csv) format, contain the wavelength of the measurement (from 400 nm to 1100 nm, at an interval of 1 nm) and the corresponding fraction of leaf reflectance, transmittance, and albedo (reflectance+transmittance). There is a separate data file for each tree and shrub species sampled at each site and a single file containing unidentified grass species collected from all of the sites. Average values for combined samples of trees and of shrubs at different sites are also provided.

This data set consists of a southern African subset of the 1-km Global Land Cover Data Set Derived from AVHRR developed at the Laboratory for Global Remote Sensing Studies (LGRSS) at the University of Maryland. Both ASCII data and binary image files are available.

This data set consists of a southern African subset of the 5-min resolution Global Potential Vegetation data set developed by Navin Ramankutty and Jon Foley at the University of Wisconsin. Data are available in both ASCII GRID and binary image file formats.