Wai‘anae Ecological Characterization

Coastal Water Quality Tool
< Back to Introduction

Example Management Application: Using N-SPECT in the Mākaha Valley

This application discusses some hypothetical examples of how N-SPECT can be applied in the Mākaha Valley. While the general application of the results are discussed, specific instructions on how to run N-SPECT are not included here. For detailed, step-by-step instructions on how to run these analyses with N-SPECT, go to the N-SPECT Tutorial, which is included in the N-SPECT installation package.

The Mākaha Valley encompasses approximately 1,800 acres of land between the Mākua Valley and the Wai‘anae Valley on the leeward coast of O‘ahu, Hawai‘i (City and County of Honolulu, Department of Planning and Permitting 2000). Together, the Mākaha Valley and the town of Mākaha are home to approximately 9,000 people (U.S. Census Bureau 2000). A variety of land uses are found in the valley, including residential areas, two 18-hole golf courses (the 300-acre Mākaha Resort Golf Club and the Mākaha Valley Country Club), a condominium and hotel development, some agriculture, and bare land. Along the coast, Mākaha Beach Park provides a place for community gatherings, recreational activities, and fishing.

Although it is difficult to predict the exact impacts that development and other land use changes will have on water quality and ecosystem health, N-SPECT is a tool that can be used to investigate the effects of land use change on pollutant and sediment loads in a watershed. N-SPECT estimates the amount of runoff, erosion, and nonpoint source pollution that occur in an area and how changes in land use could affect these loads.

N-SPECT was used in the Mākaha Valley to estimate the nitrogen and sediment loads that occur in the valley, and to investigate the effect of hypothetical land use changes on these loads. Existing land cover, soils, precipitation, and terrain data were used to estimate pollutant loads from the Mākaha Valley. Accumulated effects (which represent the amount of pollutant accumulated downstream) and local effects (which represent the source areas of nitrogen) were used as a baseline against which to compare the results when a hypothetical development or a land use delineation was incorporated. By comparing estimates of existing pollutant loads to estimates when changes to land cover/land use are incorporated, the impact of land use changes can be examined. The output from N-SPECT can also be used as an input to existing models that estimate pollutant dispersion and concentrations in the coastal ocean, based on mixing models, long shore currents, and other physical properties of the water.

It is important that the user remember that the numerical results from N-SPECT are only estimates and that there is quite a bit of uncertainty associated with these values. Therefore, exact numerical estimates and change estimates should not be relied upon as fact. The direction and magnitudes of these changes should be treated as general indicators of the direction and degree to which such management decisions impact nonpoint source pollution and erosion.

Developing an Area in the Mākaha Valley

The Department of Planning and Permitting determined that the Mākaha Valley should develop a "Special Area Plan," which serves to address issues such as cultural preservation and proposed development options (City and County of Honolulu, Department of Planning and Permitting 2000). One of the reasons for this plan was that several hundred acres of undeveloped land were zoned for residential and resort uses. The valley was also recognized as an important resource area in terms of water resources, rare and endangered plants and animals, and cultural sites. For these reasons, the valley was chosen as the area to feature in this sample application of N-SPECT.

In the Mākaha Valley, N-SPECT was used to investigate how a proposed development can affect the amount of nitrogen and erosion that occur annually, both in the developed area and throughout the watershed. An area representing the development was superimposed on to the existing land cover data and changed from its existing land cover classes to a low-intensity developed land cover class. The proposed development area is pictured in yellow. This is one of the areas in the Mākaha Valley that is specifically zoned for residential development, and this hypothetical zone was chosen due to its proximity to the golf course and the surrounding development.

Analysis of Local Effects
To examine the local effects of a land cover change, analyses of the source areas of nitrogen in the watershed were conducted, both with existing land cover and with the proposed development in place. Both analyses included identical precipitation, topographic, and soils information. The nitrogen loads from the developed area were compared to the loads from the existing land cover to estimate the direction and degree to which this hypothetical development would have on nitrogen loads.

In the figure below, the baseline nitrogen load (A) is compared to the nitrogen load with a low-intensity developed area (B). Note that, in general, the values are higher within the development polygon (higher nitrogen loads are shown in a darker shade).

When the baseline results are subtracted from the development scenario results, the areas originally classified as scrub/shrub (shown in blue) increase in nitrogen load when replaced by developed land, while the areas originally classified as grassland (shown in red) actually decrease in nitrogen load (C). The resultant changes in nitrogen relate directly back to the pollutant coefficients that are associated with each of these land cover classes, and reveal the relative contributions of these three different land cover classes to nitrogen pollutant.

Analysis of Downstream Accumulated Effects
The downstream effects of pollutants can be investigated by looking at the accumulated effects of this development. The difference in the accumulated local effects were calculated by first accumulating the results from the previous analysis (change in local effects) using the flow accumulation and flow direction data. Next, the area of the proposed development was used as a mask so that only the contributions from that area were considered in the analysis. These data sets were then used to find the accumulated load of nitrogen that is contributed by the area of interest with the existing land cover. The total percentage change with the proposed development in place was found by dividing the difference in accumulated local effects by the baseline nitrogen contributed by the area of interest. This analysis differs from the first example in that it looks at the downstream accumulated effects of the pollutants, rather than the local sources. The results provide a general idea of the direction and degree to which a management decision such as this development would have on nitrogen loads.

By looking at the point at which the watershed drains into the coastal ocean, we can investigate the impact on coastal waters. Based on this analysis, the amount of nitrogen delivered to the coast is likely to increase by approximately 63 percent. This number can be determined by querying the cell that drains into the coastal ocean. Therefore, based on this analysis, it can be concluded that this development will likely result in an overall increase in the amount of nitrogen delivered to the coastal waters.

The same process can be used to determine the magnitude of the change in sediment loads. By examining the point where the watershed drains into the coastal ocean, we can estimate the change in sediment load to coastal waters. Based on this analysis, the amount of sediment delivered to the coast is likely to increase by approximately 116 percent. This number can be determined by querying the cell that drains into the coastal ocean (shown below). Therefore, based on this analysis, it can be concluded that this development will result in an overall increase in the amount of sediment delivered to the coastal waters. This sediment load can be input into a sediment dispersion model to determine the impact on the coastal ocean environment, including coral reefs along the coast.

This example showed how the management scenario feature of N-SPECT can be used to apply a land use change to existing land cover conditions. Given the current development boom on O‘ahu, the hypothetical scenario incorporated in this exercise is a possibility in the Mākaha Valley; however, the scenario was fabricated for this example and should not be assumed to be reality. Using scenarios such as this one allows managers to better investigate the potential consequences of land use changes and their effects on nonpoint-source pollution and erosion.

Incorporating a Known Land Use in the Mākaha Valley

The previous development scenario gives an example of how a specified area can be changed to another land cover class that is already defined in the land cover classification system being used. This example will show how a land use with customized pollutant coefficients can be created and used in an N-SPECT analysis.

What can be done within N-SPECT to further refine a land cover classification system or update an outdated land cover classification? For example, the C-CAP land cover data packaged with N-SPECT were produced in 2001 (see NOAA's C-CAP Web page for more information on the data). Since then, the standard classification scheme originally used by NOAA Coastal Services Center to develop the C-CAP data was merged with the USGS National Land Cover Database (NLCD) scheme to promote consistent land cover products throughout the country. The end result of this partnership was a new classification scheme called the Coastal NLCD that added several new land cover classes to the old C-CAP scheme and, in essence, refined the original grassland and developed categories. Only 10 of the old C-CAP land cover classes are being used as the default land cover classification scheme for the Mākaha Valley. The original grassland class included both managed and natural herbaceous cover, thus lumping together areas such as meadows, pastures, and golf courses. The new Coastal NLCD scheme breaks apart these three categories into grassland/herbaceous, pasture/hay, and developed – open space classes respectively. The current version of N-SPECT with the older C-CAP data does not distinguish between these three classes, yet each class clearly impacts nonpoint-source pollution and erosion differently.

To address this issue, N-SPECT has the ability to add a land use class (including new coefficients, values for each soil type, and a cover factor) that more accurately reflects the existing land use in the area. The two 18-hole golf courses in the Mākaha Valley, which are currently categorized as grassland, can be added to the analysis with new coefficients defined specifically for golf courses.

A polygon representing the area of the golf courses, hand-digitized from high-resolution satellite imagery, was superimposed onto the existing land cover data in the Mākaha Valley. In this analysis, the area changed from its general C-CAP land cover classes to a land use defined as "golf course." The figure at left depicts the area of the golf courses.

Two N-SPECT analyses were again conducted; one with the existing land cover and one with the golf course land use incorporated. The accumulated pollutant loads from the first analysis with the existing land cover were subtracted from the results of the second analysis (incorporating the golf course) to investigate the effect that a more accurate land use classification has on the estimated nitrogen loads.

Since this area is larger than the small proposed development used in the previous example, there are several points at which the watershed drains into the coastal ocean. By looking at the changes in these values (shown below), the increase in nitrogen and sediment loading into the coastal waters can be computed. These results could then be used as input into a nitrogen or sediment model that estimates the concentrations and distribution of pollutants in the coastal environment off the Mākaha coast. An analysis could also be performed on the area in its natural state, taking it back to the land use it was prior to being developed as a golf course, to compare how pollutant levels have changed since its natural state.

This was an example of how N-SPECT outputs can change with more accurate data inputs, such as land use type. In this case, the land use feature was used to force N-SPECT to incorporate a land cover class that is not uniquely represented by the older C-CAP land cover classification scheme.

N-SPECT estimates pollutant and sediment loads from a watershed or other user-specified area and can be used to investigate fluctuations in these loads resulting from changes in land use or land cover in the area. This functionality can help users make decisions about issues in their area such as commercial and residential development (e.g., how will a new residential area affect pollutant loads?), restoration activities (e.g., will returning the landscape to its natural state reduce pollutant loads to the waters?), and resource management, both on land and in the coastal waters.

For more information on N-SPECT and how to access the tool, please see the Overview of N-SPECT. In addition to the tutorial, which walks the user through the steps to complete the examples shown here, a user's manual and technical guide are included with N-SPECT to explain the tool's interfaces and processing steps.

References Cited

City and County of Honolulu Department of Planning and Permitting. 2000. Waianae Sustainable Communities Plan. http://honoluludpp.org/planning/Waianae/Wai1.pdf

U.S. Census Bureau. 2000. Census 2000 Profiles for the City and County of Honolulu. State of Hawaii Department of Business Economic Development and Tourism Web Site. http://www.hawaii.gov/dbedt/census2k/profile-honolulu/

Back to top