John A. Harrington

Remote sensing of temporal and spatial variations in pool size, suspended sediment, turbidity, and Secchi depth in Tuttle Creek Reservoir, Kansas: 1993

Abstract Four dates of Landsat Thematic Mapper data from 1993, April 9, July 30, August 15, and September 16, were used to assess temporal and spatial patterns of lake area and dimensions of suspended sediment concentration in Tuttle Creek Reservoir, Kansas. In 1993, excessive precipitation in the Big Blue River Basin, and throughout much of the Upper Middle West, led to widespread flooding. Rains produced substantial erosion, sediment movement down the stream network, and a runoff volume that filled Tuttle Creek Reservoir, a U.S. Army Corps of Engineers flood control structure. The April 9 data are from before the flood, the July 30 data are from the time of maximum pool size and use of the emergency spillway, and the August and September data document the declining pool sizes. Three separate analyses were performed on each of the four dates of Thematic Mapper data. One set of analyses involved applying an existing physical model that uses at-satellite reflectance for TM Band 3 to estimate variations in suspended sediment, turbidity, and Secchi depth throughout the reservoir. Maps of estimated parameters of water quality for the four individual dates were compared and analyzed to document spatial and temporal changes. The second research method involved unsupervised classification (ERDAS ISODATA algorithm) of the data from the Tuttle Creek Reservoir. Water areas were grouped into coherent classes for further spatial analysis using a two-step or layered classification procedure for each date. The third analysis used a GIS overlay technique to compare the area of the water surface for each of the four dates with the flood pool as marked on U.S.G.S. 7-1/2 minute quadrangles. Comparisons document the major change in lake area between April and July, the high levels of suspended sediment in mid-summer, and the decline in pool size and concentrations of suspended sediment by mid-September. The study illustrates the advantages of using remote sensing to assist in documenting a relatively short-term environmental hazard. This study also demonstrates the value of Landsat Thematic Mapper data for use in mapping geographic variations in water area and quality in conjunction with a major flood event.

An integrated approach for water quality assessment of a Kansas watershed

An integrated modeling process was used to estimate the nutrient loadings of different sub-watersheds of the Cheney Reservoir watershed, Kansas, USA. The Agricultural Nonpoint Source Pollution (AGNPS)-ARC INFO interface was used to extract input parameters from various Geographic Information System (GIS) layers, including a landcover layer prepared from Landsat TM image, for the AGNPS model during selected storm events. The curve numbers (CNs) were adjusted depending on the antecedent moisture condition (AMC) of the sub-watershed before each storm. The storm energy intensity (EI) values were computed using a probability method from a rainfall-EI relationship for the location of the study area. Several sensitive parameters of the AGNPS model were then calibrated to match the model-estimated total nitrogen (total-N) and total phosphorous (total-P) with the measured data on a sub-watershed basis during various runoff events. This process was validated by running the calibrated AGNPS model on each sub-watershed of the Cheney Reservoir watershed. This integrated modeling process was found to be effective for smaller watersheds that had adequate rainfall data.

The Effects of Climatic Factors on Vegetation Dynamics of Tallgrass and Shortgrass Cover

Abstract Climatic factors such as precipitation and temperature place primary controls on vegetation growth and development. The relationships between these factors and vegetation dynamics, as monitored using satellite remote sensing data, have been demonstrated on continental, regional and subregional scales. Most local studies have primarily focused on small areas immediately surrounding weather stations. Here, the effects of climatic “factors over intermediate‐scale sites of relatively homogenous vegetative land cover are examined. Two areas of Kansas are selected for study: the shortgrass Sandsage Prairie in the southwest and the Tallgrass Prairie National Preserve in the east‐central Flint Hills region. Within year and between year variation in vegetation in the context of climatic variability is examined on these sites using NOAA AVHRR data. Large increases in the strength of association between vegetative response (NDVI) and one‐month lagged precipitation as compared to same month correlations are demonstrated. This does not bode well for the use of AVHRR NDVI as a “real‐time” drought monitoring tool.

Chlorophyll Measurements in the Presence of Suspended Sediment Using Broad Band Spectral Sensors Aboard Satellites

ABSTRACT Concentrations of chlorophyll-a in waters with high average annual concentrations of suspended sediment were measured in three lakes and compared with broad band Landsat MSS satellite digital data for 104 dates between December 1976 and August 1988. Concentrations of chlorophyll-a ranged from 0.3 to 211 mg·m−3 for 452 measurements. Concentrations of suspended sediment ranged from 1 to 867 mg·l−1 in the same samples. In general, chlorophyll-a concentrations were inversely related to suspended sediment concentrations. Radiance and reflectance calculated from the four MSS bands increased as a function of increasing concentrations of suspended sediment. Radiance and reflectance were inversely related to the concentration of chlorophyll-a, but the relationships were not statistically significant. This study indicated that the measurement of chlorophyll-a with broad band (100 nanometer) MSS data in waters dominated by suspended sediments is not effective since the detection of the increased absorption ...

A Climatology of Transition Season Colorado Cyclones: 1961–1990

Abstract Frequency, track, and intensity characteristics of transition season Colorado cyclones are investigated for the period 1961–90. Monthly cyclone totals are examined for evidence of seasonal frequency variations during the study period. Cyclone track maps for the months of April, May, October, and November are produced and analyzed for variations in azimuth and movement rates. In addition, the monthly distributions of minimum central pressure values are discussed. Transition season Colorado cyclone annual frequencies declined significantly during the period 1961–90, although the decline occurred primarily during the spring months. April has the most Colorado cyclone occurrences of the months examined. Spring and fall Colorado cyclone tracks are shown to be similar in trajectory and 48 h azimuth, though slightly variant in movement rates. Cyclones developing in months closer to the strong Northern Hemisphere winter circulation (April and November) regime tend to be more vigorous in terms of central ...

Vegetation change in the Mount St. Helens (U.S.A.) blast zone, 1979–1992

Abstract Mount St. Helens, Washington, erupted in May 1980 leaving thousands of acres “devastated”; by its lateral blast. Vegetation recovery has taken place through both natural processes and replanting efforts on corporate and public lands. However, an area‐wide assessment of vegetation change has been lacking. Landsat MSS data have been used to calculate NDVI for 1979 and for even years from 1980 through 1992, and to determine date‐to‐date NDVI change. The largest area showing major vegetation cover increase is Weyerhaeuser property to the northwest of the mountain. Initial analysis indicates that timing of replanting is the most important factor in determining vegetative cover increase—the faster an area was replanted after the eruption, the more recovery is evident. Three‐date images indicate rapid green‐up by 1986 in quickly replanted areas of the blast zone. Natural regeneration has been much slower, although it is evident in parts of the National Volcanic Monument.

Land Cover Change and Associated Trends in Surface Reflectivity and Vegetation Index in Southwest Kansas: 1972-1992

Abstract Land cover is an important component of the earth system. Human induced surface alteration can affect earth systems directly, through loss or degradation of ecosystems, or indirectly through impact on the climate and biogeochemical cycles necessary to sustain life on earth. The significance of the earths surface has made land use/land cover change an important issue in global change research. Alteration of land cover occurs at a variety of spatial scales, but as with many environmental change issues, the impacts of surface changes are often conceptualized at the global scale. In this study, we investigate the effects of land cover change on total reflected radiation and the Normalized Difference Vegetation Index (NDVI) in a 10,000 km2 local area in the High Plains of southwestern Kansas. Landsat MSS data from five years of record within the twenty‐year period 1973 to 1992 were classified into cool season crop, warm season crop, and pasture/prairie. Mean values of summer reflectance and NDVI from each cover type and for the study area as a whole were then analyzed for systematic change over the study period. Both reflectivity and vegetation index increased during the study period, although causes for the increase appear to be different. Results suggest that changes in mean surface reflectance in the study site are strongly influenced by land cover change, whereas changes in NDVI are more closely linked to 50‐day antecedent precipitation.

Mapping Reservoir Turbidity Patterns Using SPOT‐HRV Data

Abstract Although most orbital remote sensing systems are optimized for viewing terrestrial targets, data from these sensors has frequently been used for detecting and mapping the distribution of suspended sediments in water. In this paper, a linear regression‐based model for mapping turbidity is evaluated in a midwestern U.S. reservoir using October, 1994 SPOT‐HRV data. Results indicate that the red band (SPOT2) is the best predictor of turbidity (r=0.924, R2 = 84.5). Despite relatively low turbidity levels in the water, the radiometric resolution of the SPOT‐HRV sensor was found to be adequate for this application. The high spatial resolution of the sensor permitted observation of very detailed patterns of turbidity.

CLIMATOLOGY OF SURFACE CYCLONE TRAJECTORY AND INTENSITY FOR HEAVY-SNOW EVENTS AT THREE MIDWESTERN STATIONS

In this paper, we assess interstorm cyclone variations in intensity and motion for heavy-snow events at three north-central U.S. stations: Minneapolis, Madison, and Indianapolis. Daily data for the period from 1966 to 1996 were used to select the 157 storm events. Heavy-snow cyclone track location (latitude and longitude) and intensity (central pressure) means and standard deviations are calculated for each station and related to snowfall magnitude. The most northwesterly of the three stations, Minneapolis, had the largest median heavy-snowfall total and also the largest standard deviation in heavy-snowfall amount. Statistical analysis of surface cyclone parameters reveals significant interstation differences in storm trajectory and intensity measures for all cyclones and cyclogenesis group subsets. Chi-square tests suggest that the distribution of storms by month and cyclogenesis type is significantly nonrandom for all stations. Summary findings indicate that Midwestern heavy-snow events are dominated by southern Great Plains cyclones and that storms from secondary cyclogenic regions contribute to significant differences at the three selected locations. [Key words: cyclones, snowfall, Midwest.]

Geospatial Contributions to Watershed-Scale Surface Water Quality Modeling

First recognized in Section 208 of the 1972 Clean Water Act as a national issue, non-point source (NPS) pollution is the nation’s largest water quality problem. NPS reduction is a major challenge facing society today. The U.S. Environmental Protection Agency (EPA) estimates that agriculture is by far the single largest contributor of NPS pollution (Table 92.1), responsible for the impairment of 25 percent and 19 percent, respectively, of the nation’s surveyed river miles and lake acreage (EPA 2000).