Kenneth L. Driese

A digital land cover map of Wyoming, USA: a tool for vegetation analysis

Abstract. A Landsat Thematic Mapper (TM) based digital land cover map has been created for the state of Wyoming, USA, at moderate spatial (l-km2 minimum mapping unit) and high typal (41 land cover types) resolution as part of the Wyoming Gap Analysis Program (WGAP). This map presents opportunities for regional characterization of land cover, especially vegetation, and for examination of ecological phenomena that manifest themselves over large areas. Using the digital land cover data, we describe Wyoming vegetation and examine positions of three prominent physiognomic transitions in Wyoming: the elevation of upper and lower treeline, and the position of the biogeographic boundary between shruband grass-dominated vegetation. By area, the three leading land cover types in Wyoming are Artemisia tridentata ssp. wyomingensis sagebrush (33.4 %), mixed grass prairie (17.5 %) and Pinus contorta forest (6.5 %). Average upper-treeline elevation in Wyoming is 2947 m, and decreases with increasing latitude at an average rate of about 0.5 m/km, less than the rate of about 0.9 m/km reported by Peet (1978) for a gradient from Mexico to Canada. Lower-treeline occurs at an average elevation of 2241 m, and decreases with increasing latitude and with southerly aspect. In Wyoming, shrub-dominated communities are more likely to occur than grass-dominated communities as summer precipitation decreases below 282 mm. All of these relationships are subtle, and it appears that for particular areas, local factors are more important than regional climatic trends in explaining the position of phytogeographic boundaries.

Aerodynamic roughness parameters for semi-arid natural shrub communities of Wyoming, USA

Abstract Estimates of aerodynamic roughness length ( z 0 ) were calculated at nine sites for natural sagebrush ( Artemisia tridentata spp.), saltbush ( Atriplex nuttallii ) and greasewood ( Sarcobatus vermiculatus ) plant communities in two semi-arid basins in Wyoming, USA. Estimates were based on wind and temperature profiles measured above the plant canopies during summer (August) of 1994 and fall (September and October) of 1995. Values of z 0 were estimated for periods of near-neutral stability (fully forced convection) for four scenarios of displacement height ( d ) at each site, and for a fifth scenario based on using the low level drag coefficient for constraining friction velocity ( u ∗ ). Vegetation canopy height, fractional cover, shrub density, average canopy area (per shrub), and average biomass/area were measured for all sites. Additionally, leaf and plant area indices (LAI, PAI) were measured at three of the sites, and average shrub height was measured at the six sites. Roughness lengths calculated using an iterative method with unconstrained u ∗ averaged 0.01 m for the saltbush sites, 0.02 m for the sagebrush sites, and 0.07 m for the greasewood site. z 0 /canopy height ratios ( z 0 h c ) averaged 0.04, 0.04 and 0.13, respectively, for the three shrub types, although the average is misleading for saltbush, which showed considerable variation between sites. Roughness length appears to be related to shrub structure, as expressed by the dominant species, and by shrub density at the sites, although differences were large depending on which calculation method was used. When u ∗ was constrained, calculated z 0 and z 0 h c were smaller and less consistent in terms of relationships to vegetation structure, suggesting that further constraints on the iterations may be necessary. The results highlight the importance of improving aerodynamic roughness parameterization of natural vegetation communities.

A Vegetation Map for the Catskill Park, NY, Derived from Multi-temporal Landsat Imagery and GIS Data

Abstract A map of the vegetation of the Catskill Park, NY, was created using multi-temporal Landsat Thematic Mapper TM data and ancillary spatial data to support ecological studies in Catskill watersheds. The map emphasizes forest types defined by dominant tree species and depicts 24 vegetation classes. Mapping included a series of supervised classifications in a decision tree framework that allowed forest types to be distinguished using spectral characteristics and other environmental relationships (e.g., landscape position, elevation). Traditional contingency table analysis (based on limited ground sampling) suggests overall map accuracy ranging from 28% to 90%, depending on the level of aggregation of the original 24 map classes. Fuzzy accuracy assessment based on the same ground data suggests a 71% level of acceptable classification. The map indicates that maple-dominated forests are predominant in the Catskill region, but that beech and birch-dominated forests become more important at higher elevations. Oak-dominated forests are very important along the eastern side of the Catskills, and conifer-dominated forests are largely restricted to mountaintops and stream bottoms.

Can early season Landsat images improve locust habitat monitoring in the Amudarya River Delta of Uzbekistan

Abstract Reed (Phragmites australis) stands of the Amudarya River delta south of the Aral Sea in Uzbekistan serve as permanent breeding areas of the Asian migratory locust (Locusta migratoria migratoria). Locust swarms threaten agricultural fields adjacent to the delta. Every year, specialists from the Uzbekistan Plant Protection Service attempt to survey this vast delta to assess growth of reed which provides a habitat for locust nymphal infestations. Inferences regarding locust distribution are drawn and recommendations for chemical treatments made, based on very limited samples. This often results in blanketing wetland areas with broad-spectrum insecticides, thus harming nontarget fauna. In this study, early season Landsat data, coinciding with the locust survey planning stage, were used to generate a map of potential locust habitat. Using iterative image classification and reference data, a reed distribution map was generated with an overall accuracy of 74% (kappa agreement = 0.686). Landsat data were able to correctly identify 87% of the reed beds, but had some difficulty separating other vegetation when it was mixed with reeds. Minimizing these errors would improve the overall accuracy; however, this does not diminish the utility of this tool for locust habitat monitoring. Incorporation of remotely sensed data into current survey practices could provide precise information about the spatial distribution of reeds. Plant protection specialists could then use this to optimize planning and execution of antilocust treatments, reducing the negative environmental impact of these.

Rule‐based Integration of Remotely‐sensed Data and GIS for Land Cover Mapping in NE Costa Rica

Abstract A classification method was developed for mapping land cover in NE Costa Rica at a regional scale for spatial input to a biogeochemical model (CENTURY). To distinguish heterogeneous cover types, unsupervised classifications of Landsat Thematic Mapper data were combined with ancillary and derived data in an iterative process. Spectral classes corresponding to ground control types were segregated into a storage raster while ambiguous pixels were passed through a set of rules to the next stage of processing. Feature sets were used at each step to help sort spectral classes into land cover classes. The process enabled different feature sets to be used for different types while recognizing that spectral classification alone was not sufficient for separating cover types that were defined by heterogeneity. Spectral data included the TM reflective bands, principal components and the NDVI. Ancillary data included GIS coverages of swamp extents, banana plantation boundaries and river courses. Derived data included neighborhood variety and majority measures that captured texture. The final map depicts 18 land cover types and captures the general patterns found in the region. Some confusion still exists between closely related types such as pasture with different amounts of tree cover.

WyomingView: No-Cost Remotely Sensed Data for Geographic Education

Abstract Learning enhanced by visual examples and remotely sensed imagery is a valuable classroom resource for teaching students geographic concepts in a meaningful context. Barriers to the use of imagery include difficulty finding appropriate imagery and the cost of moderate resolution satellite imagery. A program in Wyoming called WyomingView and analogous programs in other states are providing no-cost, preprocessed satellite imagery delivered over the Internet that can help teachers better communicate geospatial knowledge to their students.

Land cover characterization for a watershed in Kenya using MODIS data and Fourier algorithms

Abstract. A time series algorithm was being presented that classifies vegetation in the Njoro Watershed, Kenya, according to the shapes of temporal normalized difference vegetation index (NDVI) profiles representing growing cycles for different vegetation. We present a two-step approach that includes noise reduction using discrete Fourier filtering and a clustering algorithm that uses the Fourier components of magnitude and phase to identify phenological differences. The classification considers possible variations in shape that may be imposed by climate, soil, topography, or human impacts. The primary input to the classification is a user-defined set of reference cycles to which pixels are assigned depending on a set of shape criteria. The output is a consistent classification of NDVI cycles representing vegetation classes with similar phenologies. The algorithm allows the creation of classification mosaics without typical boundary offsets and temporally comparable classification products. It identifies vegetation more accurately than single image classification methods, because it exploits the temporal variability in spectral reflectance due to phenological responses. We produced a classified land cover map at two hierarchical levels with five classes in a level I classification and seven classes in a level II classification that represents a refinement of the level I data. These map products compare favorably to previously published land cover maps that were developed using more standard supervised classification. The level I map has an overall accuracy of 94% compared to field data, while the level II map as an overall accuracy of 77%.