Chapter 8 - GIS and Remote Sensing Applications for Watershed

Chapter 8

GIS and Remote Sensing Applications

for Watershed Planning in the Maumee

River Basin, Ohio

Kevin Czajkowski and Patrick L. Lawrence

Abstract The Maumee River watershed is the largest drainage basin that

discharges into the Great Lakes. Although the watershed is largely a rural land-

scape, several major urban-industrial cities, including Fort Wayne and Toledo are

located along the river. Many water quality concerns are present, especially non-

point rural runoff that contributes signiﬁcant amounts of sediment into the Maumee

River. There is an important need to collect, organize and assess the available

information on the watershed conditions and to better determine the status of the

changes with land uses, crop rotation, and implementation of conservation tillage

practices within this watershed. A partnership between the University of Toledo

and US Department of Agriculture NRCS lead to several GIS and remote sensing

products including annual land cover and crop rotations via remote sensing

techniques, establishment of a Maumee Watershed Project Area GIS database,

and providing educational and informational outreach with other project partners,

resource managers, and the general public.

Keywords GIS • Remote sensing • Watershed planning

8.1 Introduction

The Maumee River watershed is the largest drainage basin that discharges into the

Great Lakes. Although the watershed is largely a rural landscape, several major

urban-industrial cities, includi ng Fort Wayne and Toledo are located along the

river. Many water quality concerns are present, including non-point rural runoff

that contributes signiﬁcant amounts of sediment into the Maumee River. There is an

important need to collect, organize and assess the available information and better

K.Czajkowski (*) • P.L. Lawrence

Department of Geography and Planning, University of Toledo, Toledo, OH 43606, USA

e-mail: [email protected]

P.L. Lawrence (ed.), Geospatial Tools for Urban Water Resources,

Geotechnologies and the Environment 7, DOI 10.1007/978-94-007-4734-0_8,

Springer Science+Business Media Dordrecht 2013

131

determine the status of the changes underway within the watershed. In 2005, the

U.S. Department of Agriculture (USDA) Natur al Resources Conservation Service

(NRCS) entered into a ﬁve year agreement with the Geographic Information

Science and Applied Geography (GISAG) Research Center at the Department of

Geography and Planning at the University of Toledo, Ohio.

The work performed assists NRCS in undertaking sub-watershed rapid resource

assessments, watershed and area planning, farm conservation planning, and delivery

of conservation technical assistance and conservation cost-share programs authorized

by the 2002 Farm Bill. The tasks undertaken with this project consists of: annually

determining land cover and crop rotations via remote sensing techniques; combining

Ohio, Indiana, and Michigan data layers to establish Maumee Watershed Project

Area GIS database; and establishing a Maumee Watershed Project GIS Website to

provide educational and informational outreach with other project partners, resource

managers, and the general public.

The Western Lake Erie Basin has been identiﬁed by NRCS as a major contributor

of non-point source pollution into Lake Erie. In 2005, NRCS developed a plan to use

Rapid Resource Assessments, Area Wide Planning, and acceleration of USDA Farm

Bill programs to address the resource concerns for the Western Basin of Lake Erie,

and contributing watersheds including the Maumee, Portage, and Ottawa Rivers.

This 10-year study primarily addresses land use/cover changes, conservation tillage

practices, and water quality monitoring. A secondary element of this plan is to deve-

lop a basin wide GIS (Geographic Information System) to aid in watershed planning

projects and public outreach. Nelson and Weschler’s (1998) study suggested that

the Maumee River watershed might not be ready for basin wide collaboration on

watershed planning, but with the implementation of a GIS-based institutional atlas,

the local and regional organizations and agencies with interests in watershed planning

could be moving in the right direction towards integration.

The watershed management approach has emerged as a holistic and integral way

of research, analysis and decision-making at a watershed scale (Montgomery et al.

1995; Perciasepe 1994; Voinov and Costanza 1999). Initially oriented towar d the

control of water supply and use, it has shifted to include a concern for water quality

and the combined effects of land use in the drainage basin, particularly since non-

point pollution has overtaken point-source pollution as a primary concern as a cause

of impairment (Nelson and Weschler 1998). By relating water quality and land use

concerns, a link is created between science and planning, thereby connecting all

stakeholders, community leaders, agency administrators, and concerned citizens

in the watershed. Basin-wide collaborations can provide the expertise, scientiﬁc

backing, moral support, and political leadership necessary to implement regional

plans. GIS interfaced hydrological models are considered as a major tool for surface

water management at a watershed scale because they are capable of presenting the

relationship between the spatial and hydrological features of the watershed in an

efﬁcient way (Al-Abed et al. 2005).

GIS is a general-purpose technology for handling geographic data in digital

form (McKinney and Cai 2002). GIS has the ability to combine physical features,

political and administrative jurisdictions, and organizational missions in order to

make sound recommendations or decisions for the entire watershed. The advances

132 K. Czajkowski and P.L. Lawrence

of GIS have grown beyond simple data management, storage, and mapping.

Today, a more sophisticated means of analysis is being utilized by combining

various mathematical and computer generated models with spatial data within the

GIS. Simulation models are useful tools for analysis of watershed processes and

their interactions, and for development and assessment of watershed management

scenarios (He 2003). Schre ier and Brown (2001) used GIS for analysis of buffer

zones, which were delinea ted and classiﬁed from digital aerial photos, which

allowed the identiﬁcation of the type, width and continuity of the buffer zone.

Kelsey et al. ( 2004) used GIS techniques to calculate several “distance or proxim-

ity” land-use variables to examine land-use effects on fecal coliform densities.

In October 2005, NRCS entered into a ﬁve year memorandum with the Geo-

graphic Information Science and Applied Geography (GISAG) Research Center

of the Department of Geography and Planning at the University of Toledo as part

of the Western Lake Erie Basin Water Resources Protection Plan. The University of

Toledo assisted NRCS in implementing the Maum ee Watershed project, including

sub watershed rapid resource assessments, watershed and area planning, on farm

conservation planning and delivery of conservation technical assistance and con-

servation cost-share programs authorized by the 2002 Farm Bill. The tasks gener-

ally consisted of: annually determining land cover and crop rotations via remote

sensing techniques; combining Ohio, Indiana, and Michigan data layers to establish

Maumee Watershed Project Area GIS data layers for the project; and establishing

and maintaining a Maumee Watershed Project GIS Website to provide educational

and informational outreach to share data and information with other project

partners, resource managers, and the general public.

Crop type classiﬁcation for the Maumee River project is bein g carried out using

multitemporal Landsat 5 satellite imagery for each year of the agreement. Images

were gathered from several time periods during the growing season to differentiate

between the different crops types, in particular corn, soybeans, wheat and pasture.

Once collected, the images underwent cloud screening and then were stacked in

Erdas IMAGINE remote sensing software package. Training sets of crop type had

been collected using a driving survey of the watershed and loca ted with Global

Positioning System (GPS) readings. These training sets were used to create spectral

signatures in Erdas and then a supervised classiﬁcation was performed using the

Maximum Likelihood classiﬁer.

8.2 Maumee River Basin

The Maumee River basin covers over 4.9 million acres across Ohio, Michigan and

Indiana. The Maumee River, the most prominent watershed in the basin, begins in

Fort Wayne, Indiana, and extends more than 130 mile s to Lake Erie, 105 miles of

which are located in Ohio (Fig. 8.1). The Maumee River has the largest drainage

area of any Great Lakes river with 8,316 square miles and drains some of the richest

farmland in Ohio. The project area for this study will only include the drainage into

8 GIS and Remote Sensing Applications for Watershed Planning in the Maumee... 133

Lake Erie south of the Ohio-Michigan state line. The cities of Toledo, Fort Wayne,

and Lima constitute the major urban areas. Other smaller towns and cities are

scattered throughout. The population of this area total s over 1.2 million people.

Land use is predominantly agriculture covering about 71% of the total basin

(NRCS 2005). Urban development and roads represent 10% of the area (NRCS

2005). Soils are naturally poorly drained. Surface ditches and subsurface drains

have been implemented to improve drainage. The basin area receives a relatively

even distribution of precipitation throughout the year between 33 and 37 in. depen-

ding on the loca tion. Soil erosion is a major problem in the basin causing NRCS

to track conservation tillage practices in order to reduce the loss of sediment off

cropland. Dredging in the Toledo Harbor, at the mouth of the Maumee River, is

costing $2.2 million per year due to sediment loading. Tourism and sport ﬁshing

are also directly related to water quality and the health of the lake associated with

increased sedimentation (NRCS 2005). Several watershed planning efforts have

been undertaken with the Maumee Basin, especially in the Maumee Great Lakes

Area of Concern (AOC) loca ted in the lower (downstream) portion of the Maumee

River and including several other rivers and streams discharging directly into the

western basin of Lake Erie (Lawrence 2003; Maumee RAP 2006 ).

Fig. 8.1 The Maumee Drainage Basin, NW Ohio

134 K. Czajkowski and P.L. Lawrence

8.3 GIS Database Development

Spatial data layers were assembled from numerous sources to assemble and deliver

GIS layers that cover the entire Maumee Watershed Project Area (Table 8.1). Some

of the websites, where spatial data is freely available, are Soil Data Mart, Data

Gateway (NRCS/USDA), Center for Geographic Information in Michigan, ODNR

GIMS (Geographic Information Management Systems), Indiana Geological Survey

(A GIS Atlas), United States Geological Survey (USGS), and Great Lakes Informa-

tion Network (GLIN). Once the data was downloaded, it was necessary to evaluate its

condition using ESRI ArcGIS software. Datasets for the basin needed to cover Ohio,

Michigan, and Indiana. In some cases, there was only Ohio datasets available and

therefore were not used. In other cases, there were security issues; therefore the data

was not made public. Metadata, which contains descriptions of the spatial data sets,

needed to be present because it indicates what has been done to create the data and

who created it. The metadata would be updated with the project purpose and contact

information. After deciding on datasets, geoprocessing techniques were performed.

Clipping, merging, and reprojecting were necessary for datasets for map overlay.

The next step was to establish a GIS website for the Maumee Watershed Project

Area linking an ArcIMS site for data viewing. The website located at www.maumee.

utoledo.edu contains background information on the project, spatial layers available

for download through a password protected ftp site, and the ArcIMS available for

viewing these spatial layers (Fig. 8.2). The ability to download the information is a

means to share the data with NRCS and project partners for the overall collaboration

of the project. Training sessions were held for partners and stakeholders to learn how

to use this online system providing them ample opportunity to make suggestions and

ask meaningful questions.

ArcIMS provides for the ability to access various layers at various scales of

viewing. In this manner it is possible to compile many important spatial data layers

into the GIS product and make each layer viewable and active at the appropriate scale.

Table 8.1 Main GIS data

layers for the Maumee GIS

Project

DEM (Digital Elevation Model)

SSURGO soils

Stream network

Land use cover

Watersheds (HUC units)

Quaternary and bedrock geology

Recreational areas and parks

Various boundaries (states, cities, and counties)

Wetlands

Source water protection areas

Groundwater data

100 year ﬂoodplains

Climatic zones

Soil drainage

Roads and transportation

8 GIS and Remote Sensing Applications for Watershed Planning in the Maumee... 135

For example, Fig. 8.3 shows the view scale of the entire watershed area highlighting

the individual river basins located within the watershed with additional layers includ-

ing land use/land cover, 2005 NIAP imagery, ecoregions and state boundaries.

Figure 8.4 illustrates a view at the scale of one river basin with display of the Land

Capability Class and additional layers that include counties, zip codes, annual precipi-

tation, farmland class and many others. Figure 8.5 displays the view at the local

community scale highlighting SSURGO soil types and also can include Census

blocks, streets, and several other additional data layers. ArcIMS also provides numer-

ous data tools to assist with spatial analysis including query functions, distance

calculations, and area measurements – all of which can be useful to potential users

of the datasets.

8.4 Remote Sensing

Land cover and land use can be classiﬁed for a watershed region by utilizing a

satellite image which covers a large area. Landsat Thematic Mapper can be used at

the regional level with its 30 m spatial resolution (Oetter et al. 2000; Jensen 2005;

Fig. 8.2 Maumee GIS project website (at www.maumee.utoledo.edu)

136 K. Czajkowski and P.L. Lawrence

Woodcock et al. 2001). The point of classiﬁcation with remote sensing is to categorize

every pixel in the image into themes or classes based on the reference spectral res-

ponse of a band. Normally, multispectral data is applied since categories which can be

separated in a channel are very limited. A commonly used method is the supervised

classiﬁcation technique which requires a prior knowledge of the study area, and pixels

are classiﬁed based on the user-deﬁned reference spectral data set. A maximum

likelihood is used to categorize the pixels into deﬁned classes as it takes into account

a variance and a covariance to the computation and classiﬁes pixels into a class to

which the pixel has the highest probability of belonging (Jensen 2005).

Many times a good classiﬁcation of land cover types can result from applying

a single image. However, when land use types such as crop types are classiﬁed,

it is useful to use multiple images wherein the dates are different. In the case of crop

identiﬁcation, the images include pre-growing season and growing season so that

different spectral information can be extracted from the images which discriminate

objects in the study area. For instance, winter wheat may be indistinguisha ble from

bare soil in late fall when it is just planted and from alfalfa in spring due to a similar

spectral response. However, by using two images, winter wheat can be identiﬁed by

having a unique set of responses to bare soil in fall and alfalfa in spring (Lillesand

et al. 2004). Therefore, it is important to know the study area to take advantage of

the multi-temporal classiﬁcat ion.

Fig. 8.3 Screenshot of Maumee Watershed layer showing all river basin units

8 GIS and Remote Sensing Applications for Watershed Planning in the Maumee... 137

A land use and land cover layer was created for this study. Throughout the

creation of the classiﬁed map, ERDAS IMAGINE 9.1 was utilized unless otherwise

noted. Ground truth points were collected at selected transits along roadways within

the watershed resulting in over 300 points annually for analysis. Landsat TM images

of path 20/row 31 were used for the classiﬁcation which were downloaded from

the OhioView website. Clouds and shadows in the images were removed by visual

assessment. Removal of the urban area was also conducted by visual assessment

using an urban area shapeﬁle downloaded from the ESRI Census Watch website

(http://www.esri.com/censuswatch). The images were then stacked to perform the

multi-temporal classiﬁcation. For example in 2005 images were used from May 4th,

August 8th, and November 12th in order to cover the complete growing season for the

primary crops: corn, soybeans and wheat.

The identiﬁcation of crop types within farm ﬁelds was also checked by USDA

personnel at the county level within the watershed on an annual basis via “wind-

shield surveys” that would generate 8,775 observation points. Among these crop

type observations, 78% were either corn or soybean. At total of 150 points per class

for the entire Maumee watershed were randomly selected as reference points, and

the others were used as training samples. Out of the 150 points, 75 points per class

Fig. 8.4 Screenshot of Maumee GIS ArcIMS product as viewing one river basin and highlighting

the land capability classiﬁcation layer

138 K. Czajkowski and P.L. Lawrence

fall into the study area and were used for accuracy assessment of the supervised

classiﬁcation for the study area. Normally, a minimum of 50 samples for each class

is good enough for the accuracy assessment. However, when a study area is larger

than one million hectare, the minimum number of the sample should be increased

to 75 or 100, thus for this study, 75 samples were used.

To perform a maximum likelihood supervised classiﬁcation, a training set for

each class of corn, soybean, hay, and wheat was created. By using the training

samples, pixels were selected by using an Area of Interest (AOI) tool for each class.

The training samples were visualized in different colors in terms of cardinal directions

to consume less effort and time in collecting pixels. For each class, about 100 ﬁelds of

pixels were collected. Those pixels were the reference for the computer to classify the

entire image. For the forest and water classes, pixels were collected visually. Water is

obvious in a satellite image by its shape and color of navy to light blue with bands 4, 3,

and 2 as red, green, and blue in color composite. Forest is also visible and easily

identiﬁable in an image by its texture and color of red with the same condition of the

color composite as that used for the water.

After running the supervised classiﬁcation, sieve and clump functions of ENVI

were applied to the classiﬁed image except for the water and forest classes to

smooth isolated pixels. The sieve function identiﬁes an isolated pixel, and the

Fig. 8.5 Screen shot of Maumee GIS ArcIMS product as viewing at the community scale and

highlighting the SSURGO soils layer

8 GIS and Remote Sensing Applications for Watershed Planning in the Maumee... 139

clump function classiﬁes the isolated pixel into the class which has the highest

occurrence of its surr ounding pixels. Some water bodies such as a river and some

forest which are represented by a pixel or line of pixels were likely to be removed

by the sieve-clump process, therefore the original water and forest classes were

reserved.

Finally, an accuracy assessment was conducted for the classiﬁed image by using

the reference points, which were separated from the training sample at the begin-

ning. The reference points were compared with the classiﬁed map to check to see if

the reference ﬁeld was classiﬁed correctly or not in ArcMap with the cardinal

directions of the refere nce point visualized in a different color. This information

was typed into Excel, which has columns of reference point numbers and classiﬁed

classes. For the forest and water class, random points were created by an accuracy

assessment function, and they were visually assessed by using the satellite image

with the color composition used for the creation of the training set described earlier

in this section.

An error matrix with columns of reference classes and rows of classiﬁed classes

was created. By using the matrix, overall accuracy was estimated by dividing the

total amount of diagonal pixels by the total amount of all of the pixels used for the

accuracy assessment. An accuracy of 85% or more overall accuracy is considered

acceptable. The accuracy of each class was estimated. A producer’s accuracy or an

omission was calculated by dividing the total number of the correctly classiﬁed

pixels in a certain class by the total number of pixels of that class derived from the

reference data. A user’s accuracy or commission errors is calculated by dividing

the total number of correctly classiﬁed pixels in a certain class by the total number

of pixels of that class derived from the classiﬁed data and tells how much the

classiﬁed pixels match with the actual validation points. A further assessment was

performed by conducting a kappa analysis, which indicates the accuracy between

the classiﬁed map and the refere nce data and if accuracy was derived from an actual

agreement between the two data or by chance. Actual agreement would be strong

with a kappa value of more than 0.80, fair with a value between 0.80 and 0.40, and

poor with a value below 0.40.

An example of an annual land cover/land use and crop type classiﬁcation (from

2005) is shown in Fig. 8.6, with farmland with planted crops the most common land

cover/land use. Soybean (47%) and corn (18%) crops are the most common rural

land use/land cover types. Forest cover was found to represent 17% of the land area.

The reference points of hay and wheat were small due to the limited amount of

the ground truth points. Overall accuracy of the classiﬁcation was 87.96% and the

Kappa value was 0.82.

Tillage classiﬁcation was conducted in the same way as the crop types were

classiﬁed. For example the tillage classiﬁcation in 2006 used a Landsat TM image

of path 20/row 31 acquired on May 23, 2006 and obtained from the OhioView

website. The classes created for the map were traditi onal tillage (<30 %), mulch-till

(30–90 %), no-till (<90 %), forest, and water. Tillage systems within the Maumee

watershed were documented at 8,927 farmﬁeld data points that were checked by

USDA personnel at the county level. Approximately ½ of the farm ﬁeld points were

140 K. Czajkowski and P.L. Lawrence

found to represent no-till agriculture in the watershed. An accuracy assessment

was also conducted, and an error matrix was created. The resulting image of the

supervised classiﬁcation conducted for the tillage system s in 2006 is shown in

Fig. 8.7. The accuracy was 81.33% and did not reach the acceptable guideline.

However, it was the best resu lt which was performed, and the technique was

utilized for the remaining annual classiﬁcation and mapping of tillage with similar

results. The Kappa value was 0.77, with the agreement between the classiﬁcation

map and the reference data fair enough and close to a strong relationship, and the

accuracy of the classiﬁcation was less likely to happen by chance.

8.5 Discussion

With the development of the Maumee GIS/Remote Sensing project several data

gathering issues have been identiﬁed. Data may have been found for only one or

two states and not complete for the entire Maumee Basin. Some contacts were

not willing to share their data due to copyright or propriety issues. Some data is

not being made public because of security reasons such as transportation, pipeline,

and other infrastructure. For many of the GI S datasets no metadata attached to data

Fig. 8.6 Land cover/land use classiﬁcation with crop types, 2005

8 GIS and Remote Sensing Applications for Watershed Planning in the Maumee... 141

sets making it very difﬁcult to validate and assess the data source and its quality.

The development of the spatial database is a continuing process. Feedback from

users of the ArcIMS web based viewer is helpful in managing the website in order

to make it more user friendly. Efforts continue in the updating of data sets with

current information.

In early 2007 a survey was sent to 188 current and potential users of the Maumee

Basin GIS/Remote Sensing project website (www.maumee.utoledo.edu) in order to

assess the utility of the website and ArcIMS data viewer and to solicit feedback on

future additions and improvements (see Rousseau, this volume for more details).

A total of 55 individuals responded (29% return rate) with state and federal

government agencies and NGOs/Universities having the largest number of resp-

onders. Over 41% of respondents indicated that they had been using GIS for more

than 5 years. When asked about their use of GIS in watershed planning 60%

indicated that they had created maps or data products and the most common sources

of GIS data that they used included the Ohio Department of Natural Resources

(ODNR) GIMS website, USGS, and NRCS Service Data Gateway. In accessing and

using the Maumee Basin GIS/Remote Sensing project website over 90% rated the

site as Good to Excellent and suggested improvements, including making more data

readily available for download, providing clearer instructions for non GIS users,

and providing links to metadata and source citation.

Tillage Systems 2006

NE Maumee Watershed

No-till

Mulch-till

Traditional tillage

Forest

Water

Urban areas/Primary roads

Watershed Boundaries

04/23/2007

5 0 10 Miles

Based on Landsat images acquired in 2006. Projected in UTM NAD 1983 Zone 17

Fig. 8.7 Classiﬁcation of Tillage Systems in the northeast portion of the Maumee Watershed, 2006

142 K. Czajkowski and P.L. Lawrence

Ongoing searches are underway to secure and post additional new useful data

sets and work continues in evaluating current data sets. Additional updates to the

data sets during the project included transet data collected annually from 2007 to

2009 on crops types for use as validation/training data for land cover analysis,

assembling data sets in response to severe ﬂooding and impacts along the Blanchard

River in August 2007, collection of the Landsat imagery for 2007–2009 for the

preparation of a series of annual land use, crop type and tillage analysis, acquisition

of one foot orthophotos for the watershed, collection of the NAIP and LiDAR

imagery available for a portion of the Maumee Basin. Tasks also included compil-

ing the ﬁeld survey data from 2007 to 2009 for crop type and tillage assessments

collected by NRCS county extension staff into a common database, analysis of

annual land cover from Landsat (including accuracy assessment and % cover

types), development of tile surface creation workﬂow to assist in creating a bare

earth DSM, and to process the LiDAR and NAIP imagery into ArcIMS .

8.6 Conclusions

The Maumee GIS/Remote Sensing Project has created an array of useful products

to assist with various watershed planning initiatives underway within the basin

including projects undertaken by the Western Lake Erie Basin Partnership, USDA

NRCS, and Army Corp of Engineers. The GIS database has been provided on

request to various parties to assist with the preparation of watershed studies, rapid

assessments, crop inventories, sediment modeling, and in response to ﬂooding

events. As further progress is made in determining the critical watershed issues

and prioritizing future projects, programs and efforts within the basin, the GIS and

remote sensing materials will be of continued importance. The proj ect website, with

GIS data layers and the land use/land cover mapping from remote sensing, provides

many important opportunities for public outreach, teaching and the creation of

unique map based products for a variety of watershed projects within the Maumee

Basin. With the ongoing development of watershed based planning within the

Maumee Basin by a variety of agenc ies and organizations these spatial data sources

will be increased value in plan development.

Acknowledgements Work undertaken under this project has been funded by a Memorandum of

Understanding between the University of Toledo Department of Geography and Planning GISAG

Research Center and the USDA National Resource Conservation Service (NRCS) for 2005–2010.

Appreciation is extended to Steve Davis, Cheryl Rice and NRCS staff for their assistance.

Dr. Kevin Czajkowski and Dr. Patrick L. Lawrence with the Department of Geography and

Planning served as the project principal investigators. James Coss and Tim Ault provided technical

assistance as research associates. Graduate students from the MA Geography program at the

University of Toledo: David Dean, Katie Swartz, Phil Haney, and Rumiko Hayase completed

the GIS and remote sensing components.

8 GIS and Remote Sensing Applications for Watershed Planning in the Maumee... 143

References

Al-Abed N, Abdulla F, Khayarah AA (2005) GIS hydrological models for managing water

resources in the Zarqa River basin. Environ Geol 47:405–411

He C (2003) Integration of geographic information systems and simulation model for watershed

management. Environ Model Softw 18:809–813

Jensen JR (2005) Introductory digital image processing: a remote sensing perspective, 3rd edn.

Pearson Prentice Hall, Upper Saddle River

Kelsey H, Porter DE, Scott G, Neet M, White D (2004) Using geographic information systems and

regression analysis to evaluate relationships between land use and fecal coliform bacterial

pollution. J Exp Mar Biol Ecol 298:197–209

Lawrence PL (2003) Science and the management of the Great Lakes coast: case study of the

Maumee Area of Concern. In: Proceedings of the 2003 Canadian coastal conference. Canadian

Coastal Science and Engineering Association. Queens University, Kingston, Ontario

Lillesand TM, Kiefer RW, Chipman JW (2004) Remote sensing and image interpretation, 5th edn.

Wiley, New York

Maumee RAP (2006) Maumee area of concern stage 2 watershed restoration plan (Draft). Ohio

EPA Northwest District Ofﬁce, Bowling Green www.maumeerap.org

McKinney DC, Cai X (2002) Linking GIS and water resources management models: an object-

oriented method. Environ Model Softw 17:413–425

Montgomery DR, Grant GE, Sullivan K (1995) Watershed analysis as a framework for

implementing ecosystem management. Water Res Bull 31:369–386

Natural Resources Conservation Service (2005) Western Lake Erie Basin water resources protec-

tion plan: Ohio, Indiana and Michigan. From Natural Resources Conservation Service. http://

www.oh.nrcs.usda.gov/programs/erie_basin/erie_basin_project.html. Retrieved 8 March 2007

Nelson LS, Weschler LF (1998) Institutional readiness for integrated watershed management: the

case of the Maumee River. Soc Sci J 35(4):565–576

Oettter DR, Cohen WB, Berterretche M, Maiersperger TK, Kennedy RE (2000) Land cover

mapping in an agricultural setting using multiseasonal Thematic Mapper data. Remote Sens

Environ 76:139–155

Perciasepe R (1994) EPAs watershed approach to ecosystem management. Fisheries 19:4

Schreier H, Brown S (2001) Scaling issues in watersheds assessments. Water Policy 3:475–489

Voinov A, Costanza R (1999) Watershed management and the Web. J Environ Manage

56:231–245

Woodcock C, Macomber SA, Pax-Lenny M, Cohen WB (2001) Monitoring large areas for forest

change using landsat: generalization across space, time and land sensors. Remote Sens Environ

78:194–203

144 K. Czajkowski and P.L. Lawrence