A Method for Understanding the Hydrologic Implication of Comprehensive Land Use Plans

Renee Gunn, Jane Frankenberger, Bernie Engel

Agricultural and Biological Engineering Department, Purdue University

Introduction

The population of Indiana has increased 9.7% since 1990, with many Indiana families moving into subdivisions built on past agricultural lands. This increased development and resulting land use changes have a significant impact on both water quality and water quantity. As communities grow, local officials generally concentrate on the economic implications of growth (i.e., tax base, number of jobs, new housing starts, etc.) and pay little attention to the water resource implications of this growth. Land use plans (and resulting planning and zoning activities) must consider the impacts of potential land use changes on water resources if water quality goals are to be achieved. However, there are few models for how this can be done.

Two simple indicators of water resource impacts resulting from development are runoff depths determined by the NRCS Curve Number Method (USDA, 1986) and percent impervious surface. The NRCS Curve Number Method was developed as a simple, easy to use method to determine runoff from an area based on rainfall, land use and hydrologic soil group. The basis of the method is the Curve Number, which is determined by the land use and the hydrologic soil group. The low data requirements and the ease of use has made the Curve Number Method a very popular method, and it is used in many more complex models. The L-THIA (Long-Term Hydrologic Impact Assessment, http://www.ecn.purdue.edu/runoff) model, which is based on the NRCS Curve Number Method, was used in this analysis to determine runoff depths. The percent impervious surface of a watershed has been shown to have a direct effect of the quality of the water and streams in the watershed (Scheuler, 1994). Impervious surfaces are any area in which water cannot infiltrate into the ground such as rooftops, roads, parking lots and sidewalks. The majority of impervious surfaces are in the form of roads and parking lots.

One objective of this study was to analyze the hydrologic implications of three land use scenarios in Howard County, Indiana; 1) the current land uses (Current Land Use) 2) a full build-out based on the current comprehensive land use plan ('Worst Case Build-out) 3) a full build-out based on a proposed land use plan (Proposed Planned Growth). A secondary objective was to document the process used in this study so that others may conduct their own analysis with a comparable approach.

Study Area

The unincorporated area of Howard County, Indiana was chosen for this example analysis based on three factors; the large rate of growth around a metropolitan area, the similarity of Howard County to other Indiana counties and the availability of data. The areas immediately outside the city limits of Kokomo, the major metropolitan area in Howard County, are experiencing a high rate of development, mainly residential. This type of growth, residential and small amounts of commercial outside the city limits, is not limited to the large cities in Indiana, but also many small to medium size communities. The current zoning in many of these counties either does not restrict the subdivision of land, or the land is being rezoned at a rapid rate.

In 1997, Howard County Plan Commission, which has jurisdiction in the unincorporated area of the county, saw the need for a new comprehensive land use plan for the county and hired a consulting firm to create a new comprehensive plan with the input from the community. Comprehensive plans usually contain information on where the county is, where they want to go, what it will look like and how to get there. The important information for this project in the Proposed Howard County Comprehensive Plan includes current zoning maps and explanation of the zones, current and expected future sewer districts, and the Proposed Planned Growth zoning maps.

To better understand the effects of the comprehensive plan on the water resources of Howard County an analysis of possible future changes in land use and their long-term effect of the hydrology was conducted using L-THIA. The Current Land Use was used as a base line to compare the impact of water resources of future land uses. The 'Worst Case Build-Out' is based on the maximum amount of development allowed under the current zoning. A full build out of the proposed land use plan is the basis of the Proposed Planned Growth analysis.

Methods

Data Sets

The first step in the analysis was the organization of the data. Soils data for the county and three sets of land use or zoning data were used in this project: Current Land Use, current zoning and Proposed Planned Growth zoning. Current Land Use The Current Land Use (Figure 1.) is based on GAP land use data. (ISURSL, http://baby.indstate.edu/geo/rs/ ) . This data, derived from satellite imaging, was originally created to determine natural cover types for vertebrates to help define 'gaps' in the habitat. In order to simplify this analysis, many of the non-urban categories were combined in order to reduce the original 18 classifications of land use to 6. The categories used in this project are explained in Table 1.

Figure 1. Current Land Use of Unincorporated Howard County, Indiana

Current Zoning Paper maps of the current zoning were available from the Proposed Howard County Comprehensive Plan (Bernardin et al., 1998). The maps were scanned and imported into ArcView. The ArcView Image Analysis extension was used to rectify the scanned image to its correct geographic location using an available 'roads' layer as a reference. The images were stretched and moved until the roads on the image lined up with the roads layer. The zoning classes were then screen digitized using functionality within ArcView.

Figure 2. Current Zoning of Unincorporated Howard County, Indiana

'Proposed Planned Growth' The new Comprehensive Plan for Howard County lays out a plan for future growth within the county. This Proposed Planned Growth Zoning (Figure 3) recognized that the majority of the residential growth would be near the cities and towns, while farmland away from town needs some protection. This plan allows higher density residential (1/2 to 1/4 acre lots) subdivisions near the city, sets aside an area for 'country homes' on 1/2 to 2 acres in an outer ring and increases the minimum lot size far away from the cities to 2 acres to promote agriculture in the area. An expanded commercial district was also planned on the north edge of the city. By promoting growth near the cities, better planning can be done for services such as roads, sewers, schools and fire stations.

Figure 3. Proposed Planned Growth of Unincorporated Howard County, Indiana

Hydrologic Soil Group Another set of data needed for this project was the NRCS Hydrologic Soil Groups. This data is complied by the NRCS as part of the soil surveys for each county. The data used in this analysis is from the STATSGO data set. This data is available for the entire country in digital form, but is not as detailed as the SSURGO data that is available for several counties in Indiana, but currently not Howard County. The hydrologic soil group is an indication of the runoff potential of the soil. 'A' type soils, such as sandy soil, have very low runoff potential . Heavy clay and mucky soils are of type 'D' and have very high runoff potential. The digitized soil map for Howard specified only B and C type soils that are the most common in northern Indiana.

Map 4. Hydrologic Soil Group of Unincorporated Howard County, Indiana

Curve Numbers and Percent Impervious Surface

The Curve Number was found for every land use and hydrologic soil group combination in the three scenarios; Present Land Use, 'Worst Case Build-out', and Planned Proposed Growth. Many of the land use and hydrologic soil group combination were determined from the NRCS publication Urban Hydrology for Small Watersheds, TR-55 (USDA, 1986) while others were determined for a specific land use not listed. The modifications of the Curve Numbers resulted in non-whole numbers. The listing of Curve Numbers to the tenth is for comparison of changes to the Curve Numbers when factors such as conservation tillage or roads are included. When Curve Numbers are used in the equations they are always whole Numbers as directed in the USDA publication TR-55 (USDA, 1986) and are used as such in this project.

Percent impervious surface was also found for every land use used in this project. Some land uses like agricultural, forest and grassland had a very low percent impervious surface due to the scattered houses and few roads in the area. Other more urban land uses had much higher percent impervious surfaces. The percent impervious surface for residential areas was modified from the percentage listed in TR-55 to account for roadways.

Current Land Use

The majority of unincorporated Howard County is being used for agriculture. There are areas of forest mainly along the banks and in the floodplain of Wildcat and Kokomo Creeks. Most of the residential housing is on the south and west sides of the city of Kokomo with only a small amount of Industrial or Commercial on the northern side.

The Curve Number and percent impervious areas had to be determined for all of the land use/ soil group combinations in this scenario. The Curve Numbers for the land uses; Forest, Grass/pasture, High Density Urban and Water, were found directly from TR-55. The High Density Urban classification from the GAP classification indicates Commercial/Industrial and thus, the Curve Number for Commercial was used. For agriculture and Low Density Urban, the Curve Numbers could not be directly determined from TR-55.

The Curve Numbers representing agricultural land in Howard County were found by using a weighted average of the various Curve Numbers for each land use based on the acreage of each crop or tillage type (Table 3.). For agricultural areas the Curve Number varies by crop and tillage type. The data was not available for each field in Howard County, but the acreage of fields under different cropping or tillage practices was.

The major crops in Howard County are corn and soybeans and to a lesser degree wheat and hay. The average acreage over the past 10 years for each crop was found using the NASS (National Agricultural Statistics Service) on-line database (http://www.nass.usda.gov:81/ipedb/) and is summarized in Table 3. Conservation tillage practices such as no-till have long been known to reduce runoff from agricultural fields. 80% of the soybeans and 17% of the corn in Howard County in 2000 was planted using conservation tillage practices (Evans et al., 2000). The acreage of corn and soybeans were adjusted to reflect the approximate amount of conservation tillage in Howard County.

The Curve Number and the percent impervious surface of the urban land uses were based on the lot layout and the road requirement. TR-55 reports that a 1/2 acre residential lot is 25% impervious, but this does not include the roadway required in front of the house. Based on the roadway width, right-of-way and minimum lot width required by Howard County, the minimum amount of roadway required for a lot was determined (See Appendix B. Calculations) First, the area needed in front of the lot for the right-of-way and road was added to the lot area. The road was considered 100% impervious with a Curve Number of 98, the remaining area of the right-of-way was considered grass with a percent impervious of 0% and a Curve Number of 61 or 74 depending on the hydrologic soil group. The total lot Curve Number and percent impervious surface was found from computing a weighted average, based on acreage, of the impervious surfaces and the grass (Table 5).

The minimum lot width, right-of-way and roadway width were not given for commercial areas. Since the percent impervious surfaces were already quite high, they were assumed to not change greatly if road and right-of-ways were taken into account.

Even in land classified as agriculture or grass/pasture, there are some impervious areas. For the rural areas, the impervious surfaces were mainly roads, which were not separated from the agricultural areas in the GAP classification. Based on the road survey in the Howard County Comprehensive Plan, 2800 acres, approximately 2% of the land classified as agriculture is impervious (Table 3). Aerial photos showed that portion of the area classified as grass/pasture in the GAP data was actually Low Density Residential or Farmstead. (Table 5)

'Worst Case Build-Out'

The 'Worst Case Build-out' scenario demonstrated what could happen to land use patterns in Howard County based on its current zoning, although it is not considered likely. This included building houses on 1/4 acre lots and roads to service the houses throughout the current residential areas (zoned R1), expanding current industrial and commercial areas (zoned B1, B2, C1 or I1) to the extent of the zoning and converting all of the area currently zoned A1 to residential subdivisions with 1/2 acre lots and the roads needed to access these areas.

The area currently zoned R1 was treated the same as the low density urban from the Current Land Use classification. There is a stipulation in the Howard County Zoning ordinances concerning lot size based on whether the residential area is serviced by sewers. If the area is not serviced, the minimum lot size is 1/2 acre compared to 1/4 acre for serviced areas. The sewer service district map in the Howard County Comprehensive Plan was compared with the area zoned R1. There was only a small area on the east side of town that was zoned residential and not currently in the sewer district. Therefore it was assumed that there were no zoned residential areas without sewers. Residential subdivisions with half acre lots were 'built' in all of the area zoned A1. Also included were the minimum acreage of roads required for these subdivisions (See Appendix B. Calculations). By including the roadway both the Curve Numbers and the percent impervious surface are increased from the values listed in TR-55. The Curve Number and percent impervious surface for this area were computed similarly to the low density residential land use from the present land use scenario.

Map 5. 'Worst Case Build Out' Land Use

All of the area zoned A1 was assumed to have residential subdivisions built upon it in accordance with the current zoning regulations. Current zoning regulations state that permitted land uses include agriculture and single-family dwelling at an intensity of 2.18 dwelling units per acre. As with the area zoned R1, the required minimum roadways were included in the calculations of Curve Number and percent impervious surfaces (Appendix B. Calculations). The Curve Number and percent impervious surface was computed similarly to the low density residential land use from the present land use scenario. This resulted in the addition of about 1/4 million houses on 1/2 acre lots in the area currently zoned A1. Not only did this add quite a few rooftops to the area it also added a large number of roads.

The Airport Hazard Zone can only be developed into urban areas by special exception, so this area remains agriculture. Although agriculture is allowed in the Flood Plain District according to the zoning ordinances, some of this area is currently forested and will most likely remain in that state. Because of this the Flood Plain District is assumed to be 50% agriculture and 50% forest. The Curve Numbers and percent impervious surfaces for agricultural areas and those used in calculations for the Flood Plain District are identical to those used in the Current Land Use scenario.

The final residential zoning included on the maps is the Very Low Density. The large lot size is to discourage subdivision growth and to keep the land in agricultural production or forested. For this reason, the Curve Number and percent impervious surfaces are based on current agricultural practices, farmsteads, a few residential houses and roads (Appendix A.2. Very Low Density Residential).

This plan does not include a different zoning classification for the flood plain; some of this area is currently forested and will most likely remain in that state. Because of this, the Flood Plain District has been included in this scenario and is assumed to be 50% agriculture and 50% forest and the Curve Number and percent impervious surface are identical to the Flood Plain District in the 'Worst Case Build-out'. The plan also incorporated an expanded commercial district on the north side of town that will include a new automotive plant in the near future. The Curve Number and percent impervious surface used for the commercial district are those found in TR-55.

Average Annual Runoff

To determine the average annual runoff using the NRCS Curve Number method, L-THIA (Long-Term Hydrologic Impact Assessment) was used. L-THIA is a straightforward analysis tool that provides estimates of changes in surface runoff and nonpoint source point pollution (NPS pollution) resulting from past or proposed land use changes. It estimates long-term average annual runoff for land use and soil combinations, based on actual long-term climate data for that area. By using many years of climate data in the analysis, L-THIA focuses on the average impact, rather than an extreme year or storm.

Because the data for this project was compiled and created in terms of GIS maps, and the GIS version of L-THIA was used. The GIS version automates the process of impact modeling within the ArcView GIS. The model produces maps of runoff depth and volume along with nonpoint source pollution loading of the area. The process used in this project with the GIS version of L-THIA is relatively straight forward and follows the steps laid out in the publication L-THIA NPS available at the L-THIA website (http://www.ecn.purdue.edu/runoff click on L-THIA GIS). Using custom Curve Numbers is covered in Appendix II and used in this project. Only the runoff portion of L-THIA was used in this project, but determining the nonpoint source pollution values is a simple next step.

Results

Curve Number and Percent Impervious Surface

Current Land Use

'Worst Case Build Out'

Proposed Planned Growth Land Use

In the Current Land use map there is a similarity between agriculture and low density urban Curve Numbers, but a large difference in the percent impervious surface. Even with the Curve Number for agriculture being decreased for use of conservation tillage in the county and Curve Number for low density urban being increased due to the inclusion of roads in the subdivision, low density urban still has slightly lower Curve Numbers than agriculture. The effects of tillage on compaction and the disruption of the soil profile increased the Curve Number for agriculture. In contrast, the Curve Number for low density urban is based on the large amount of impervious surfaces from rooftops, driveways and roadways and the large amount of grassed lawns. The percent impervious surface in agriculture, forest and grass/pasture are a result of houses, barns and roads within these areas. By comparing the areas classified as agriculture, forest and grass/pasture with the aerial photographs it was seen that there were scattered houses and farmsteads in these areas. This was especially evident in the grass/pasture area where a large portion of the area was actually large acre residential lots.

Although differences in Curve Numbers between 1/2 acre lots and 1/4 acre lots is small in the 'Worst Case Build Out' land use, the resulting change in runoff will be of greater magnitude due to the non-linearity of the Curve Number methods. Based on the required width of the road and minimum lot width, between 6% and 9% of the new ½ acre subdivisions will need to be roads, not including new or upgraded major thoroughfares. These roadways, along with the rooftops, driveways, etc. will increase the percent impervious surface for both of the residential areas above the 25% percent impervious surface limit that Scheuler states results in degraded streams (Schueler, 1994).

Runoff

The NRCS Curve Number method was designed for "comparing the effectiveness of alternative systems of measures within a watershed project. [It is] not used to recreate specific features of an actual storm."(Kent, 1966) Because of this, runoff results from different land uses are compared to each other rather then focusing on specific values. The runoff values for the three scenarios are shown above with values listed for a reference on the magnitude of the change.

The values for the Current Land Use scenario are mainly in the low to medium range. Although there are very few impervious surfaces, the agricultural fields still produce runoff. This is mainly from soil disturbance due to tillage, lack of ground cover and compaction from farm machinery. The areas of B hydrologic group soil have slightly less runoff than the C soils. In this county, the majority of the B type soils occur near Wildcat Creek or other smaller creeks.

When houses and roads are added in the 'Worst Case Build-out' scenario, the runoff remains decreases slightly. Although the percent impervious surface increases from 2% to 29% from the addition of rooftops, driveways, and roads, there are large areas of lawn also added. The NRCS Curve Number Method assumes that the lawns act similarly to open grassland or pasture which has high infiltration rates and thus lower Curve Numbers. Limited research with conflicting results has been conducted in this field, so the assumption that lawns act similar to grassland was used in this project.

In the Proposed Planned Growth Scenario, the ring nearest the cities and towns had the highest runoff rates for residential areas. These rates decreased as the number of houses per acre decreases. The very slight difference in the runoff between the three residential zones in the Proposed Planned Growth land use shows that the hydrologic soil group has more of an effect on the Curve Number than the lot size in the residential areas. This suggests that when planning for future growth both the proximity to cities and the hydrologic soil group should be considered. In all of the scenarios, the commercial and industrial areas have the highest amounts of runoff, mainly due to the large amount of impervious surface on site such as parking lots and roofs.

Summary and Conclusions

By using the L-THIA model three land use scenarios, Current Land Use, 'Worst Case Build-out' and Proposed Planned Growth were compared for the unincorporated areas of Howard County, Indiana. The majority of work in this type of project is the compilation and development of the data sets. The data, the steps and assumptions used to develop the data and the analysis using L-THIA are outlined in this paper. Development of data included classifying the Current Land Use data into categories that L-THIA uses, making assumptions and calculation, regarding roads, and interpreting the Proposed Planned Growth Zoning. Converting land use into Curve Number and percent impervious surface was the final step in developing the data. The L-THIA model was then used to compare runoff from the three scenarios.

By comparing the three scenarios it was shown that by allowing unrestricted growth in the area currently zoned A1, the runoff would slightly decrease, but the percent impervious area will be above 25%. By adopting the new comprehensive plan the county can focus residential growth near Kokomo and the smaller towns, thus persevering the agricultural land and reducing the hydrologic impact of the growth on the county as a whole.

This paper presents a method for understanding the effects that comprehensive plans can have on the hydrology of the county by either not being restrictive as in the current plan for Howard County or by planning for growth as the proposed plan does. The results of this project also include Curve Number and percent impervious surface values for typical Midwest land use that are not included in TR-55.

The methodology outlined in this paper can be used to analyze past, present or future land uses in other areas. Other analyses can use the Curve Numbers or percent impervious surface values that were found in this project, or use the methodology presented to determine Curve Number and percent impervious surface for another land use.

Bibliography

Bernardin, Lochmueller and Associates, Inc, 1998. Comprehensive Plan for Howard County, Indiana. Bernardin, Lochmueller & Associates, 6200 Vogel Road, Evansville, Indiana, 47715-4006.

Central Massachusetts Regional Planning Commission, Build-Out Analysis: Sources, Assumptions and Methodologies. http://www.cmrpc.org

Evans, M.G., K.J. Eck, B. Gauck, J.M. Krejci, J.E. Lake and E.A. Matzat, 2000. Conservation Tillage Update: Keeping Soil Covered and Water Clean in the New Millennium. Purdue Cooperative Extension Service, Department of Agronomy. AGRY-00-02.

Kent, K.M., 1966. Estimating runoff from rainfall in small watersheds. Presented at meeting of Am. Geophys. Union, Los Angeles, California.

IBRC, 1999. Indiana's Population Growth Continues. http://www.stats.indiana.edu/population_menu.html

National Agricultural Statistics Service. http://www.usda.gov/nass/

Scheuler, 1994. The Importance of Imperviousness. Watershed Protection Techniques, Vol1, No. 3 - Fall 1994

U.S. Department of Agriculture Soil Conservation Service, 1986. Urban Hydrology for Small Watersheds. Technical Release 55, US. Dept of Agriculture, Washington, D.C.

Appendix A. Tables

Appendix B. Calculations

Curve Number and Percent Impervious Residential Subdivisions

(Modified from Central Massachusetts Region Planning Commission)

http://www.ci.grafton.ma.us/planning/buildout/buildout-methodology.html

Calculations shown for 1/2 acre lots HSG B for example:

The Curve Number and percent impervious area is a weighted average of the Curve Numbers in the drawing above. Note, only ½ of the road and right-of-way are used (14 ft and 11 ft respectively); the other half is for the house across the street.