<p> An important aspect in spatial analysis is a digital separation technique. There are several techniques possible, but trend-residual techniques and filtering techniques are emphasized in this report. The areas studied were a 9.6 km^2 gridded area in the Mammoth Cave region in Kentucky and the Lewisburg region in West Virginia. The basic geology and surface patterns are described. Although they are similar in respect to geology, the doline distribution is much smaller and more dense in Kentucky than in West Virginia. Both areas appear stable and subsidence dolines appear dominantly. The two doline development models are presented and the evidence cited tends to support the MDCP model. The predicted clustering of daughters about uniform/randomly distributed parents is found in Kentucky, no study was found for West Virginia. Both areas show that doline long axis are oriented parallel to regional joint sets.</p> <p> A uniformly dense grid proves accurate and unbiased when the elevation data are contoured. The Surface II Graphics package proves more than adequate in producing all the maps needed for this study despite the lack of flexibility in some areas.</p> <p> A trend-residual analysis was conducted to the fourth order for both the Kentucky and West Virginia areas. Despite minor technical problems, the results are positive. The method clearly separates the local doline variance from the regional trend, but there does appear to be a consistent bias towards nearby ridges. It also appears that there is an enhancement-suppression effect from the residual analysis such that certain doline forms are enhanced while others are suppressed as the trend order progresses from the first to the fourth order.</p> <p> The filtering technique also shows some excellent results from the digital separation method. Several filter types are discussed and the theory of their design is also presented. Very successful results were achieved by a zero summed filter as well as a 3 x 3 moving average filter. Several maps are produced from the analyses and are computer generated and these are also presented. There are several conclusions given at the end.</p> / Thesis / Candidate in Philosophy

This research enthusiastically highlights the bio-adsorption of methylene blue (MB) by local, poultry, NaOH and citric acid modified ubiquitous eggshell (LES, NLES, CLES, PES, NPES and CPES) adsorbents. The microstructures of these adsorbents indicated that they had some surface functional moieties that were responsible for the adsorption of MB. The Langmuir isotherm and PSO model best fit the experiment data. The largest Langmuir monolayer adsorption capacity qmax, was 242.47mg/g, with the largest MB initial concentration of 400mg/L. This was a clear indication and a confirmation that MB adsorption by the powdered eggshells was chemisorptive. Moreover, the values of F $F$, the thickness of the boundary layer/film were >0, showing that the rate limiting step for the adsorption process was controlled by more than one diffusion mechanism. The values of ΔG° for the adsorption of MB by the adsorbents indicated that the adsorption reactions were all non-feasible and non-spontaneous. The values for ΔS° (J/K/mol) for LES, NLES and CPES for the uptake of MB showed decrease in the chaos or degree of randomness of the adsorption reactions, and the reverse was the case for PES, NPES and CLES for the uptake of MB, which showed increase in the chaos or degree of randomness of the adsorption. The adsorption of MB by LES, NLES and CPES gave ΔH°(kJ/mol) values which were indicative of endothermic nature of the adsorption systems, and the reverse was the case for the uptake of MB by PES, NPES and CLES, which was indicative of the exothermic nature of the adsorption systems.