The study area is part of the glaciated region of northeastern Kansas. Unconsolidated surficial sediments include Pleistocene till and loess on uplands and Holocene alluvium in the stream valleys (fig. 4). Underlying bedrock is composed of alternating layers of limestone, sandstone, and shale of Pennsylvanian age. The high ridge in the western portion of the military reservation is supported by the Oread Limestone of the Shawnee Group. The Oread includes thick limestone members that are relatively resistant to erosion. Lower portions of the study area are underlain by softer shale and sandstone of the Douglas Group. The upland study forest is associated mainly with the Knox and Gosport soil complexes (fig. 5), which developed from loess, till, and weathered bedrock. They are located on moderately to steeply sloping portions of the topographic ridge in the western portion of the military reservation. The bottomland study forest grows mainly on Onawa soils in alluvial sediment of the floodplain.
Fig. 4. Generalized geology of the study area. Based on data (1999) from DASC, Kansas Geological Survey.
Fig. 5. Soils map of the study area. Based on data (1999) from DASC, Kansas Geological Survey.
The climate of northeastern Kansas is continental with distinct seasons. Mean annual temperature is about 54° F (12° C); temperature extremes range from more than 100° F (38° C) to less than 0° F (-18° C). Daily temperature fluctuations of 35° F (20° C) or more can happen any time during the year. Average annual precipitation is around 90 cm (35 inches). Most precipitation is received in spring and early summer (April-July) with a second wet season in the autumn (October-November) of some years. The mean annual precipitation (January-December) for the study area for the period 1987-1997 was 74 cm (29 inches), well below the long-term average (NCDC, 1999).
Photosynthetically active land vegetation strongly absorbs red (0.6-0.7 µm) light and strongly reflects near-infrared (0.7-1.0 µm) energy. It is the only common land-cover material with this spectral signature (Colwell, 1974). Many vegetation indices have been derived based on numerical combinations of red and near-infrared values in remotely sensed datasets (Tucker, 1979; Perry and Lautenschlager, 1984). The most widely used vegetation index is the Normalized Difference Vegetation Index (NDVI), introduced by Rouse et al. (1974) and now a standard method for vegetation analysis: NVDI = Near-infrared - Red / Near-infrared + Red (Price and Bausch, 1995).
NDVI values range from -1 to +1. Positive values represent active vegetation, and near-zero or negative values represent other types of materials: bare soil/sediment, bedrock, concrete/asphalt, water bodies, snow/ice, and clouds. Early work revealed that plants under stress show reduced reflectivity in the infrared spectrum (Gates et al., 1965; Allen and Richardson, 1968). Further research has shown that moisture content in plant leaves is indicated by the ratio of mid-infrared (1.3-2.5 µm) to near-infrared reflectivity (Tucker, 1980; Carter, 1991). On this basis, a moisture-stress index (MSI) has been utilized to detect drought and other stress conditions in vegetation and is intended to detect decreased water content in leaf foliage: MSI = Mid-infrared / Near-infrared (Rock et al., 1986; Hunt and Rock, 1989; Cibula et al., 1992).
In a preliminary analysis of Landsat TM datasets for the Fort Leavenworth upland forest (Wilkins, 1997), NDVI values for spring, summer, and autumn for 1987, 1988, 1990, and 1994 were examined and compared to climatic records and tree-ring data. Wilkins (1997) identified a one- to two-year lag between the drought of 1988-1989 and decline in forest NDVI values. In another study based on AVHRR data, Wang et al. (2001) examined biweekly trends in NDVI, temperature and precipitation during a 9-year period (1989-1997) for the state of Kansas. They found that deviations in average precipitation corresponded to most spatial and year-to-year variations in NDVI values; however, temperature deviations from average generally were not correlated with NDVI variations. They concluded that precipitation is a strong indicator for regional patterns of NDVI and vegetation productivity.
Dendroclimatology is based on analysis of tree-ring records. Variations in tree-ring width are highly correlated with availability of moisture in soils (Fritts, 1976), and dendroclimatology has been studied in several parts of the U.S. Great Plains (Stockton and Meko, 1975; Blasing and Duvick, 1984; Cook et al., 1988). In an investigation of the tree-ring record for oak species in the upland forest at Fort Leavenworth, Nang (1998) found that tree-ring width is correlated strongly with the Palmer Drought Severity Index (PDSI), a common climatic index for drought conditions (Palmer, 1965), but not strongly correlated with either precipitation or temperature alone. Nang (1998) concluded that growth of tree rings responds in the same year to available moisture. NDVI values derived from Landsat TM datasets, however, do not show a direct relationship to tree-ring width or PDSI values (Nang, 1998).
In similar studies elsewhere, Davenport and Nicholson (1990) utilized NDVI values to analyze various types of vegetation in East Africa. They found that NDVI values are sensitive to rainfall in dry areas, but a significant lag period was noted for wet forests. In the southeastern United States, Pinder and McLeod (1999) found rather weak correlations among decreased rainfall, tree-ring growth, and MSI values for longleaf pine. These results taken collectively indicate that complicated relationships exist between climatic events, tree growth, forest canopy, and NDVI values determined from satellite imagery. Climatic events may result in immediate responses for some tree-growth factors, whereas responses may be delayed by one or two years for other factors. Furthermore, the nature of forest response to climatic events may vary with different tree species and environmental settings.
Kansas Geological Survey
Web version January 25, 2002
http://www.kgs.ku.edu/Current/2002/aber/aber3.html
email:lbrosius@kgs.ku.edu