Watershed vulnerability was assessed for Bernalillo County, New Mexico using a multi-criteria Fuzzy Inference System (FIS) implemented in a Geographic Information System (GIS). A vulnerability map was produced by means of a weighted overlay analysis that combined soil erosion and infiltration maps derived from the FIS methodology. Five vulnerability classes were stipulated in the model: not vulnerable (N), slightly vulnerable (SV), moderately vulnerable (MV), highly vulnerable (HV), and extremely vulnerable (EV). The results indicate that about 88% of the study area is susceptible to slight (SV) to moderate vulnerability (MV), with 11% of the area subject to experience high or extreme vulnerability (HV/EV). For land use and land cover (LULC) classifications, shrub land was identified to experience the most vulnerability. Weighted overlay output compared similarly with the results predicted by Revised Universal Soil Loss Equation (RUSLE) model with the exception of the not vulnerable (N) class. The eastern portion of the county was identified as most vulnerable due to its high slope and high precipitation. Herein, structural stormwater control measures (SCMs) may be viable for managing runoff and sediment transport offsite. This multi-criteria FIS/GIS approach can provide useful information to guide decision makers in selection of suitable structural and non-structural SCMs for the arid Southwest.
Urban sprawl is a leading cause of non-point source (NPS) pollution in the world [
To address this concern, scientists, planners, and stakeholders are usually required to evaluate several competing watershed-specific factors in order to develop procedures and practices that can best minimize the issues associated with stormwater runoff. Common watershed factors include topography, precipitation, soil, vegetation, and drainage area. In most cases, selecting the best combination of methods and alternative conservation practices in the presence of competing factors poses a great challenge for decision makers [
Structural SCMs and non-structural SCMs involve techniques, methods, activities, and practices that have the primary objective of preventing or reducing the quantity of pollutants present in runoff water before reaching natural water systems. The increase in SCM awareness has been partly attributed to their capability to control runoff volume while preserving existing hydrological function [
One extensively used technique for improving SCM efficiency is by incorporating multi-criteria decision support (MCDS) systems. MCDS involves the use of scientific and mathematical tools to evaluate complex trade-offs between different environmental and socio-economic factors [
This study combines specific soil erosion and infiltration criteria within a GIS-based FIS framework using Fuzzycell software to develop a watershed vulnerability map for Bernalillo County, New Mexico. The area is characterized by intermittent precipitation and limited water availability. Model output could be beneficial for exploring structural and non-structural SCM approaches to watershed management for targeted vulnerable areas within the county. The premise herein is that areas with high soil erosion potential and low infiltration potential are most vulnerable. An AHP based vulnerability map was developed concurrently [
Irrespective of the wide acceptance of SCM approaches in stormwater control projects, there has been little representation of such approaches in the desert Southwest [
Use of economic argument to justify the expense and burden of SCM projects is a difficult task. One reason for this perception can be attributed to the fact that direct economic benefits of SCM projects are not readily quantifiable. Contrary to this belief, a study of the economic benefits of conservation practices in five states (Florida, Nebraska, New Mexico, Maine and Oregon) indicated a direct relationship between economic growth and watershed health [
Fuzzy Inference Systems (FIS)
Fuzzy inference systems (FIS) uses fuzzy set theory to classify data based on the degree of memberships. In classical set theory, an element can either exist as a member of a set (1) or a non-member (0). Contrary to classical set theory, fuzzy set theory classifies data based on the degree of memberships ranging from 0 to 1 (
Numerous attempts have been made to investigate the significance of fuzzy logic in watershed studies. Previous approaches, such as that by Nielsen and Hjelmfelt [
In recent years, there has been an increasing amount of literature on fuzzy based water conservation techniques. One example of such approaches is that by Shamsudin et al. [
harvesting system as an alternative source of potable water. A more innovative example has been presented by Ki and Ray [
Watershed vulnerability, defined herein as high soil erosion and low infiltration potential, is a complex process to model. Several watershed factors such as slope gradient, length slope factor, soil texture, soil erodibility, soil permeability, landuse and landcover, vegetation cover, and drainage density influence the development of the watershed vulnerability model. The characteristics and effect of these factors on soil erosion and infiltration are outlined in this section.
Slope Gradient (S)
Slope gradient (S), usually expressed as percent slope or degree slope, is used to represent the degree of variability in elevation and is found to strongly influence the velocity and direction of stormwater runoff [
Length Slope Factor (LS)
The length slope factor (LS) is an important topographic factor used in the Revised Universal Soil Loss Equation (RUSLE) computations to predict long term soil loss risks [
Soil Texture
Soil texture, classified with respect to percentages of sand and clay, is known to have a strong influence on the rate of infiltration and runoff potential of watersheds [
Soil Erodibility (K)
The soil erodibility (K) factor is a quantitative soil property that plays a major role in determining the extent to which soil particles are eroded or transported by runoff water. Parameters such soil moisture, soil permeability, and organic matter content influence the degree of soil erodibility and stability. Soils with lower K values are highly stable compared to those with higher K values. The K factor is also an important parameter in the RUSLE soil erosion prediction calculations. Values for the K factor are available via the NRCS soil map STATSGO [
Vegetation Cover
Normalized Difference Vegetation Index (NDVI) is a widely used standardization parameter to represent vegetation cover and land surface. NDVI values typically range from +1.0 to −1.0 with non-vegetated areas such as streams and rivers producing small or slightly negative values, and vegetated areas producing values ranging from 0.4 to 1.0 [
Soil Permeability (Ksat)
Soil permeability, or more correctly saturated hydraulic conductivity (Ksat), is an important physical soil property that refers to the ability of soils to conduct water. Ksat values give an indication of the degree of perviousness or imperviousness of soils. High Ksat values represent the portions of watersheds that are highly drained and are suitable for infiltration practices. Conversely, lower Ksat values correspond to watershed regions that experience high runoff and are prone to soil erosion.
Landuse and Landcover (LULC)
Landcover refers to the physical materials such as forest, crops, grass, and water covering the earth’s surface, while landuse refers to the use of land for human activities. LULC directly translates to the amount of water penetration into soil layers or being produced as runoff water. For example vegetative surfaces act as a protective layer which slows down runoff and allows adequate detention time for infiltration and groundwater recharge to occur [
Precipitation
Area precipitation characteristics (type, duration, intensity, and amount) are highly important in drainage projects due to their direct influence on the volume of generated runoff. Precipitation that falls on impervious surfaces eventually is carried to natural drainage. On porous surfaces, however, precipitated water may serve as a significant source of recharge for groundwater aquifer zones and rainwater harvesting sites. Precipitation characteristics play a significant role in soil erosion. Precipitation intensity and duration on soil erosion are important factors. In semi-arid watersheds, high intensity precipitation storms are believed to be main component for soil erosion. However, an event based erosion model applied to a catchment in southeast Spain indicated that large magnitude, low frequency events potentially contribute significantly to total soil erosion [
Drainage Density
Drainage density is a quantitative indicator that represents the ratio of the total length of streams to the total watershed area. Low drainage density values are favorable for infiltration while larger values indicate high runoff potential. Recent findings indicate that adequately drained watersheds have drainage densities less than 1 km/km2 while poorly drained watersheds have drainage densities greater than 5 km/km2 [
The overall modelling approach used herein is based on using a MCDS FIS approach (
Soil Erosion Sub-Model
Soil erosion is considered as one of the leading causes of environmental degradation in the world, contributing to long-term poor land productivity, surface and groundwater contamination, and loss of valuable natural resources. Nearly 99% of the total suspended solids (TSS) found in rivers and streams can be attributed to soil erosion in the United States [
Infiltration Sub-Model
In order to fully address water availability concerns, it important to understand the impact of infiltration in water conservation programs. About 48% of irrigation water and 98% of domestic water in the United States are derived from groundwater [
To examine infiltration potential within a MCDS FIS framework, attributes such as soil characteristics, vegetative cover, slope gradient, drainage density, and land-use are needed as input. For example, soil hydraulic conductivity significantly influences infiltration, as does vegetative cover with respect to surface entry of water. Slope gradient and drainage density affect surface runoff, which impacts overall infiltration. Thus, these interrelated factors may be used to assess infiltration potential using a MCDS FIS approach.
Soil Erosion Infiltration Potential (SIP) Model
Watershed protection programs should not only focus on soil erosion, but also infiltration as it is an integral component of the watershed surface and groundwater hydrology. Soil erosion potential is reduced with increased infiltration as overall surface runoff decreases. An innovative approach to watershed management is to develop a watershed vulnerability model that incorporates soil erosion and infiltration models. Mosadeghi et al. [
This study was conducted using maps and digital data from Bernalillo County, New Mexico. The county has a total area of about 3023 km2 (1167 mi2) and is centrally located at longitude 106.669 and latitude 35.054. Having a population density of 220 persons/km2 (571 persons/mi2), Bernalillo County is the most populous county in the State of New Mexico, largely due to the City of Albuquerque. The region is covered by shrubs and scrubs (71%), deciduous, mixed, and evergreen forests (18%), associated urban (9%), and open water (1%). The two dominant soil orders are Entisols (47%) and Aridisols (32%). Bernalillo County is selected for this study because it falls within urban growth and high residential areas within New Mexico. Much of the region has an elevation of more than 1400 m. The 100 year 24 hour (100P24) storm event ranges from 96 to 164 mm (2.45 to 4.16 in) across the area. Limited water availability, unique soil characteristics, dominant shrubland vegetation, and intricate climate, coupled with a forecasted 35% population growth over the next 20 years, makes Bernalillo County a suitable case study area for the current research.
Tools and Data
A variety of software tools and databases are needed to facilitate implementation of an FIS approach to assess watershed vulnerability. The following sections summarize the tools and databases used for this research, as well as the extensive data preparation and processing required.
ArcGIS®: A Geographic Information System (GIS) provides an extensive toolset to manage multiple layers of spatially-distributed geographic data. Modelling approaches based on GIS have been shown to be successful in assessing regional scale soil erosion risks [
TauDEM: Terrain Analysis Using Digital Elevation Models (TauDEM) are a collection of Digital Elevation Model (DEM) processing tools [
Soil Data Viewer: The NRCS Soil Data Viewer is an ArcGIS® extension for analyzing soil data and building soil-based thematic maps. The program enables map makers to interpret soil attributes and properties without having to deal with the complexity associated with querying and processing soil databases. The Soil Data Viewer v 6.2 was used to create thematic layers such as soil texture, hydrologic soil groups, soil depth, and saturated hydraulic conductivity (Ksat) maps for the study area.
Thematic Layers: Several factors such as slope, soil erodibility, precipitation, LULC, soil type and texture, and vegetation were used in developing the watershed vulnerability model. The majority of the thematic layers were derived from base map layers such as LULC, DEM, soil data, and national hydrographic dataset (NHD) maps. The LULC map was provided by the National Land Cover Database [
Data Preparation and Processing: A grid size of 30 × 30 m pixel was chosen for all spatial layers to cover the study area extent. All acquired data with cell size different than 30 m pixel size were geometrically corrected using the cubic resampling technique in the Data Management toolbox of ArcGIS®. The Data management toolbox was then utilized to clip the input thematic layers to conform to the spatial extent of the study area. The Define Projection tool was used to project maps to the appropriate coordinate system in order for the maps to line up and display correctly in ArcGIS®. All layers were projected using the NAD 1983 Albers projection and the North American 1983 datum. In order to produce the soil texture and Ksat maps, the soil map together with the soil survey database (SSURGO) was loaded into the NRCS’s Soil Data Viewer program. Fuzzy membership functions, rule generation, and defuzzification of crisp inputs were conducted using the Fuzzy Rule Builder and FuzzyCell programs. In order for FuzzyCell to function properly, all input maps for the program were converted to the GRID raster format. All final map manipulations and overlay analysis were completed using the ARCGIS10 platform.
Derived Model Attributes
For the watershed vulnerability analysis herein, multiple objective criteria may be derived using the ArcGIS® environment.
Slope Gradient: A slope gradient map (degrees) was extracted from a digital elevation model (DEM) of the study area using the Slope tool in ArcGIS®. The resulting slope gradient map revealed that the eastern portions of the study area had high slope gradient values, while slope values varied from low to moderate towards the west.
D-Infinity Flow Accumulation: A flow accumulation (flowacc) map is needed to determine LS factor values for the RUSLE application. The flowacc map was created based on a three-step procedure within the TauDEM environment. The first step involved the development of a depressionless DEM raster using the Pit Remove tool. The purpose of creating a depressionless DEM raster was to ensure that areas enclosed by higher elevation values are filled in order to avoid erroneous flow-direction computations. Secondly, the D-Infinity Flow-Direction tool was then used to create a flow direction raster layer. The final flowacc raster was created using the depressionless DEM and the flow-direction maps by means of the D-Infinity Contributing Area tool in TauDEM.
Length Slope Factor (LS): The LS map for the study area was generated according to the overland flow method [
Drainage Density: The Line Density function in Spatial Analyst toolbox of ArcGIS® was used to extract drainage density values from a hydrographic dataset. Drainage density values for the study area ranged from 0 to 25.54 km/km2.
Support Practice Factor (P): The support practice factor (P) value thresholds were determined based on the slope gradient (S) map expressed as percentages as proposed by Ahamed et al. [
Development of FIS Model
Fuzzy logic is a decision support tool that contributes to a better representation and modeling of GIS-based data [
FuzzyCell: FuzzyCell is among the most widely used fuzzy logic tools for processing geographic data. It allows users to incorporate human knowledge spatial analyses, classify data into membership functions, generate “if-then” rule aggregation, and provide inference methods for defuzzification. FuzzyCell was used as the main fuzzy inference system for classifying GIS raster data in fuzzy memberships and defuzzifying inputs. All generated fuzzy output maps are based on the Mamdani inference system and the Center of Area (COA) defuzzification method.
Fuzzy Rule Builder (FRB: Fuzzy Rule Builder (FRB) is a custom fuzzy inference tool developed in Microsoft Visual Studio 2010® express in order to handle fuzzy standardization and generating “if-then” fuzzy rule aggregation. The fundamental concept behind fuzzy inference processing involves the construction of a collection of “if-then” rules. The role of the “if then” rules are to establish a framework for converting crisp input data into defuzzified outputs. The total number of rules generated was determined as Fn, where n is the total number of input factors and F is the number of membership functions. A total of 243 fuzzy rules were generated for each model. Rules from the program were subsequently exported to FuzzyCell environment for fuzzy inference and defuzzification.
Fuzzy Membership Classifications for Watershed Vulnerability
Each input variable for fuzzy logic processing was coded into 3 membership classes (low, moderate, and high). The relationships between each input parameter and the output risk maps were used to establish membership classification threshold values. The trapezoidal shaped membership function (with a partial overlap to account for uncertainty) was used as the basis of establishing membership classes. Trapezoidal membership function curves were chosen due to its unique potential to optimize the performance of fuzzy inference processing [
Fuzzy Standardization and Weighting
To be able to utilize input layers of variable spatial representation in fuzzy analysis, it was important that layers be associated with a common scale of measurement. This process of transforming input layers to a common scale (0 - 1) is termed standardization. All fuzzy input variables were transformed into a standardized input layer either using Equation (1) or Equation (2) for positive and negative correlations, respectively, with soil erosion or infiltration thematic layer inputs.
w i = x i − x i min x i max − x i min (1)
Thematic Layer | Low | Medium | High |
---|---|---|---|
Slope (˚) | 0 - 5 | 5 - 8 | 8 - 72.37 |
100P24 (in) | 2.45 - 2.70 | 3.0 - 3.40 | 3.70 - 4.16 |
K-Factor | 0 - 0.1 | 0.2 - 0.3 | 0.4 - 0.55 |
NDVI | 160 - 145 | 145 - 125 | 125 - 115 |
Landuse | Wetlands, Forest, Cultivated, and Water | Shrub | Urban, Barren |
Thematic Layer | Low | Medium | High |
---|---|---|---|
Drainage Density (km/km2) | 0 - 1 | 1 - 5 | >5 |
Ksat (μm/s) | 141.14 - 10 | 10 - 1 | <1 |
Soil | A | B, C | A/D, C/D, D |
Slope (˚) | 0 - 5 | 5 - 8 | 8 - 72.37 |
Landuse | Wetlands, Forest, Cultivated, and Water | Scrubs | Urban, Barren |
w i = x i max − x i x i max − x i min (2)
where wi is a standardized weight, xi is an initial value of criteria i, xi max is the maximum value, and xi min is the minimum value [
The development of the watershed vulnerability model presented in this study involved 3 major steps: 1) preparation of thematic layers, 2) extraction of fuzzy memberships using Fuzzy Cell, and 3) overlaying FIS results to produce the watershed output map. The watershed model depended on the determination of soil erosion and infiltration potential and identifying factors involved. Two sub-models were created including a soil erosion model and an infiltration model.
Vulnerability Map: The final watershed vulnerability map is a 30 × 30 m cell raster produced by overlaying soil erosion and infiltration potential maps obtained from the FIS process. The overlay process was conducted in ArcGIS® using the Weighted Overlay Tool. Equal weight was assigned to each map in the overlay analysis. A detailed illustration of the vulnerability assessment procedure is given in
Model Comparison with RUSLE: Empirical soil erosion risk prediction models such as RUSLE continue to play a vital role in soil conservation and protection programs. RUSLE is a universally accepted empirical method developed by United States Agricultural Research Service [
E = R K L S C P (3)
where E (tons/ac/yr = 2.242 tonnes/ha/yr) is the predicted average annual soil loss, R (hundreds of ft.tonf.in/ac/hr/yr) is the rainfall erosivity factor, K (ton/ac/unit R) is the soil erodibility factor, LS is the length slope factor, C is the cover management factor, and P is the support practice factor [
The soil erosion map by merging five input raster layers, as defined in Equation (3). In order to compare the results to the watershed vulnerability model, the RUSLE output was subsequently divided into five vulnerability groups in tons/ac/yr, where 1 ton/ac/yr equals 2.242 tonnes/ha/yr: 1) not vulnerable (0 - 2); 2) slightly vulnerable (2 - 6); 3) moderately vulnerable (6 - 50); 4) highly vulnerable (50 - 150); and 5) extremely vulnerable (>150). These groups span across what is considered sustainable soil loss at 2 - 5 ton/ha/yr [
(r2) between the two maps.
Evaluation of Watershed Vulnerability
The assessment of the watershed vulnerability for this study was performed in ArcGIS® through a weighted overlay analysis of infiltration and soil erosion potential maps. The soil erosion maps were developed as a result of merging 5 input raster data: slope gradient, precipitation, K factor, NDVI, and LULC. Similarly, the infiltration map was developed by merging drainage density, soil permeability, soil texture, slope gradient, and LULC raster layers. The final output of each approach and subsequent overlay, which comprised an overall vulnerability map, is given in
As displayed in
Evaluation of Soil Erosion and Infiltration Risks
In this study, the distributions of soil erosion and infiltration risks were also considered. In order to compare the model outcomes, the five vulnerability classes (N, SV, MV, HV, and EV) used for the watershed vulnerability model were applied to both the soil erosion and infiltration risk models. The risk distribution of soil erosion and infiltration models followed a somewhat different pattern relative to the combined watershed vulnerability model, with soil erosion showing a higher percentage in the slightly vulnerable category and infiltration having a higher percentage in the moderately vulnerable as did the combined overlay.
Suitability Rating | N | SV | MV | HV | EV |
---|---|---|---|---|---|
RUSLE | 29.52 | 28.83 | 33.94 | 6.06 | 1.37 |
Infiltration FIS | 1.47 | 19.96 | 55.42 | 22.47 | 0.62 |
Soil Erosion FIS | 4.75 | 61.65 | 23.94 | 8.59 | 1.01 |
Watershed FIS | 0.65 | 12.09 | 76.01 | 9.87 | 1.38 |
N = Not Vulnerable, SV = Slightly Vulnerable, MV = Moderately Vulnerable, HV = Highly Vulnerable, and EV = Extremely Vulnerable.
Infiltration risk followed a similar pattern; however, the infiltration risk distribution also revealed high risk areas around the central and western portions of the study area. The vulnerability distributions for soil erosion and infiltration risk are summarized in
Comparison with the RUSLE Model
The areas of watershed vulnerability for the study area were contrasted against soil erosion prediction from the RUSLE model. The spatial distribution of soil erosion vulnerability for Bernalillo County is given in
LULC | N | SV | MV | HV | EV | |
---|---|---|---|---|---|---|
11 | Water | 0 | 0.23 | 0.03 | 0.001 | 0 |
21 | Urban | 0.02 | 1.75 | 6.91 | 0.053 | 0.012 |
31 | Barren | 0.001 | 0.04 | 1.41 | 0.060 | 0.011 |
41 | Forest | 0.69 | 2.71 | 9.84 | 6.12 | 0.32 |
51 | Shrub | 0.04 | 5.96 | 33.87 | 1.96 | 0.75 |
81 | Cultivated | 0.007 | 5.16 | 20.83 | 0.66 | 0.12 |
91 | Wetlands | 0 | 0.17 | 0.24 | 0.001 | 0 |
N = Not Vulnerable, SV = Slightly Vulnerable, MV = Moderately Vulnerable, HV = Highly Vulnerable, and EV = Extremely Vulnerable.
Influence of Landuse and Landcover (LULC)
An attempt was made to quantify the influence of landcover and landuse activities on the watershed risk assessment using the 7 LULC classes given in
The results suggest that shrub, cultivated land, and forest zones are the most affected zones. About 65% and 9% of these LULC areas fell in the moderately vulnerable and highly vulnerable category, respectively, with slightly over 1% classed in the extremely vulnerable category. In general shrub lands were found to be the most vulnerable LULC category (highest total percentage in the top three vulnerability classes). This aligns with expectations based on area climate and literature. Ground cover, or vegetative litter, often may be sparse in semi-arid shrub lands. The role of this layer in modulating surface runoff and soil erosion remains poorly understood; however, in general, runoff volume and sediment yield are reduced in the presence of litter cover [
Use of Average Annual Maximum NDVI
NDVI input was a criterion for the soil erosion sub-model only. Hydrological partitioning at the watershed scale is a function of vegetative cover. Seasonal precipitation in New Mexico is distinctly bimodal with less intense spring rains and late summer monsoonal rains. As such a bimodal NDVI seasonal cycle is observed with a late spring-early summer initial peak, a mid-late summer pre-monsoon dip, and a second maximum peak in early autumn [
Multi-Criteria Decision Support approaches such as FIS provide a relatively easy way of determining watershed vulnerability distributions on a regional scale. From the viewpoint of decision makers, these methods can provide several strategic benefits including the ability to incorporate project objectives and a variety of quantitative data. One possible explanation for its gain in popularity is the lack of statistically meaningful data and methodology to deal with issues associated with uncertainty. Overlapping memberships addresses some uncertainty in the FIS approach.
As demonstrated in this study, FIS can be an effective means of generating regional scale vulnerability map for watershed investigations. The FIS approach integrated both soil erosion potential and infiltration potential using specific criteria to characterize watershed vulnerability, whereas the RUSLE model only accounts for soil erosion. The FIS approach offers several benefits in developing regional scale watershed vulnerability model relative to conventional field surveys of watershed vulnerability, in terms of time and effort and degree of complexity. It is important to acknowledge, however, that detailed field investigations might be required in order to extend the applicability of the methodology discussed herein.
Quantifying a regional scale risk index map can provide spatially relevant information for engineering management and decision making. In view of this, a 30 × 30 m watershed vulnerability risk map was generated for the Bernalillo County, New Mexico using an FIS methodology. The approach incorporated drainage density, slope, precipitation, vegetative cover, landuse and landcover, soil erodibility, soil texture, and soil permeability as the primary factors contributing to watershed vulnerability. The final map was produced by merging soil erosion and infiltration potential using the Weighted Overlay method in ArcGIS®.
Analysis of the vulnerability map identified that the most problematic areas were present in the eastern portions of the study area. Based on the analysis, less than 12% of the study area was identified to fall within the high vulnerability (HV) or extreme vulnerability (EV) category, whereas less than 1% of the total area was associated with no vulnerability (N). Around 88% of the area was predicted to experience slight (SV) to moderate (MV) vulnerability. About 9% of shrubs, cultivated lands, and forest zones were predicted to have high vulnerability (HV) with shrub land being most vulnerable class overall.
In terms of spatial extent and general patterns, the FIS soil erosion vulnerability output map visually compared with the RUSLE prediction. The AHP approach showed a higher correlation with RUSLE than the FIS method [
The current lack of high resolution spatial data and current site-specific landuse and land cover (LULC) information may limit the potential to field-test the watershed vulnerability model. It is, therefore, important that future extensions of the model focus on generating high resolution (e.g. a 10 × 10 m raster data) especially for the highly vulnerable regions and also obtaining detailed site-specific LULC information. NDVI data appropriate to the critical pre-monsoonal period should be developed to better address soil erosion potential. In addition, photographic documentation from various field sites is needed to understand how the landscape looks for different classes of soil erosion risk and infiltration potential. In order to improve on the functionality and applicability of the model, other factors such public awareness and economic indicators could be incorporated into the model.
Information from the FIS watershed vulnerability model can be useful for managing new structural or non-structural SCM projects or improving existing ones to mitigate impact to stormwater quality and quantity for identified vulnerable areas. A recent study showed that runoff water originating from nearby urbanized watersheds in Albuquerque, NM carried high levels of polychlorinated biphenyls (PCBs) into the Rio Grande [
The authors declare no conflicts of interest regarding the publication of this paper.
Richardson, C.P. and Amankwatia, K. (2019) Assessing Watershed Vulnerability in Bernalillo County, New Mexico Using GIS-Based Fuzzy Inference. Journal of Water Resource and Protection, 11, 99-121. https://doi.org/10.4236/jwarp.2019.112007