Conrad D. Heatwole

TMDL Modeling of Fecal Coliform Bacteria with HSPF

Fecal coliform TMDLs were developed for nine watersheds in Virginia using the HSPF model. The primary HSPF algorithms used to simulate FC loading and fate in the models are described in detail. Parameter values are summarized for all HSPF parameters related to FC simulation, as well as source data used external to the model for developing input loads from the various individual FC sources. Although there are many areas of uncertainty in modeling fecal coliform, a scientific approach was used in the evaluation of sources, the representation of the sources, and the evaluation of parameters used to simulate fecal coliform fate and transport with the HSPF model for nine sub-watersheds. The similarity of source reductions called for in each of the nine TMDLs support recommendations for the regional application of key results from TMDL studies to watersheds with similar sources and for the use of adaptive implementation as presented in a recent National Research Council report to Congress assessing the scientific basis of TMDLs (NRC, 2001).

Assessment of interpolation methods for refining SRTM and DEM

Shuttle Radar Topographic Mission (SRTM) DEM is widely used in analyzing topography of developing countries because it is freely available. But due to its coarse resolution (3 arc-second or 90m), the SRTM is sometimes not appropriate for use in hydrologic analysis and deriving stream networks. The purpose of this study was to evaluate the potential and performance of seven interpolation techniques as a means of refining the resolution to 30m. IDW with exponents of 0.5, 1, and 2, regularized and tension Spline, Kriging, and Natural Neighbor interpolation were tested using ArcGIS 9.1 Spatial Analyst. The obtained elevation, slope, aspect, sink, and stream networks from the interpolated 30m DEM were compared quantitatively with those derived from the original 30m SRTM DEM. The Renasco River area of New Mexico was selected for the analysis area because it has a range of topographic features from flat lands and hills, to mountains.

Representing fuzzy, uncertain evidences having variable weights: confidence propagation for rule-based systems

Abstract Representing knowledge uncertainty, aggregating evidence confidences, and propagating uncertainties are three key elements that affect the ability of a rule-based expert system to represent domains with uncertainty. Fuzzy set theory provides a good mathematical tool for representing the vagueness associated with a variable when, as the condition of a rule, it only partially corresponds to the input data. However, the aggregation of and ed and or ed confidences is not as simple as the intersection and union operators defined for fuzzy set memberships. There is, in fact, a certain degree of compensation that occurs when an expert aggregates confidences associated with compound evidence. Further, experts often consider individual evidences to be of varying importance, or weight, in their support for a conclusion. This paper presents a flexible approach for evaluating evidence and conclusion confidences. Evidences may be represented as fuzzy or nonfuzzy variables with an associated degree of certainty. Different weights can also be assigned to the individual conditions in determining the confidence of compound evidence. Conclusion confidence is calculated using a modified approach combining the evidence confidence and a rule strength. The techniques developed offer a flexible framework for representing knowledge and propagating uncertainties. This framework has the potential to reflect human aggregation of uncertain information more accurately than simple minimum and maximum operators do.

Assessment of SWAT Model Uncertainty of Soil Parameters for Agricultural Dominated Systems under Data-Poor Conditions

Data-poor environments represent places where there are not enough data to accurately characterize and evaluate watershed systems. Available data are often scarce, incomplete or nonexistent, consequently are poor-quality and often uncertain. One-phase Monte Carlo Simulation methodology was used to characterize the uncertainty in SWAT’s outputs due to input parameters likely to be lacking in data-poor conditions. The analysis was performed for two experimental watersheds, Little River, GA and Reynolds Creek, ID. Uncertainty due to selected soil parameters was quantified as a function of DEM resolution (10, 30 and 90m) and soils database (SSURGO and STATSGO). Uncertainty was calculated for a 12-year period for streamflow, sediment yield, total nitrogen and total phosphorous. Changing the soils database resulted in a significant (p < 0.001) change in model predictions. A change in DEM resolution did not always impact the predictions; the results depended on the type of watershed, constituent of interest, and soil parameter. The combined effect (Soil_DB x DEM) on model predictions was less than the effect of each factor. In general, predictions had lower values and increased uncertainty as the input data got coarser, but there were some exceptions. The SSURGO soils database combined with the 30-m DEM appears to be an adequate level of resolution. Therefore, additional research should be focused on improving existing information in data-poor environments to represent similar resolution to SSURGO and 30-m DEM.

A simple distributed overland and channel routing method for the Time-Area approach to develop direct runoff hydrograph

The time-area approach provides a useful tool to understand runoff mechanisms and to construct hydrographs through its intuitive and inherently distributed concept. Some researchers have applied this approach to develop unit hydrographs (UH) at outlets. However, the UH concept has limitations in application due to assumptions of a uniform distribution of excess rainfall and a constant base time of the direct runoff hydrograph regardless of excess rainfall intensity. Therefore it may not be appropriate to use a UH in distributed hydrologic modeling of spatially varied excess rainfall. A direct method is presented to derive hydrographs from the distributed excess rainfall map using the time-area map (isochrone) without the assumptions of UH.

Trends and Patterns in Luangwa River Valley Fires 2003 ~ 2007

Efforts at improving resource management, sustainability, and wildlife habitat in the Luangwa River Valley of Zambia includes community education to change the common practices of setting fires throughout the long dry season. Fires are set for many reasons -- to clear fields of last years crop residue, to clear new land, to remove undergrowth for safety along roads and paths, accidentally, and sometimes apparently with little reason.