Zachary Carango 1. Modeling Soil Loss Due to Erosion in Northwest Maui. Problem

Transcription

1 Zachary Carango LA221 Modeling Soil Loss Due to Erosion in Northwest Maui Problem Erosion and related soil loss in northwestern Maui poses a significant threat to fragile coral reefs nearby. When washed into the ocean at an anomalously high rate, sediment can overwhelm and kill coral, irreversibly damaging ecosystems. The cause of increased soil loss in Northwestern Maui is not currently known although it is suspected that human activity, whether through changes in land use, the introduction of nonnative species, or climate change is ultimately responsible. Understanding patterns of erosion and runoff are critical to mitigating coral reef damage in the short term and developing a long term solution to soil runoff. I used a the Fastscape landscape evolution model along with data about soil type, vegetation, and land use to determine which areas likely contribute most to sediment loss. It is my hope that this model will help planners and land managers work towards a solution for protecting and ensuring the long term health of Maui s coral reefs. History Soil loss in tropical environments such as Maui has been studied for decades (El Swaify et al, 1982), although a thorough effort to model erosion in northwestern Maui has not yet been published. The Universal Soil Loss Equation is a simple, empirical relationship that predicts the rate of soil loss based on the following factors: rainfall, soil erodibility, slope, slope length, vegetation (Weesies et al, 1997). While the universal soil loss equation serves as a good estimate for erosion, it does almost nothing to account for topography and drainage networks. These features, which reflect the shape of the landscape, can greatly affect differences in the erosion rate at small spatial scales. Because of this, accounting for topographic influences on

2 2 erosion is critical to creating a well targeted and cost effective solution for soil loss in Northwestern Maui. Fastscape is based on the stream power law, an alternate formula for modeling erosion which does a much better job of accounting for topography (Braun and Willett, 2013). The stream power law simplifies factors such as soil quality and vegetation but does a far better job accounting for slope and the flow of water than the universal soil loss equation. The stream power law determines erosion at a location based on the area drained through that location. In other words, the stream power law is able to model the entire drainage network of an area with reasonable accuracy. Both the universal soil loss equation and the stream power law have strengths the other lacks. I believe that by incorporating aspects of the universal soil loss equations with a Fastscape like implementation of the stream power law I can make confident predictions about which areas are most prone to erosion and most suited to soil loss prevention efforts in northwestern Maui. Methods My analysis mathematical methodology is based on the Universal Soil Loss Equation and the stream power law. The universal soil loss equation is a largely empirical relationship that combines many qualitative factors with slope and slope length (Figure 1). Despite its relatively simple form the universal soil loss equation has been used for many years to predict erosion with reasonable success. KLSCP A = R Figure 1. Universal soil loss equation.

3 3 The universal soil loss equation incorporates the following information (Weesies et al.): Rainfall (R), Soil Erodibility (K), Slope (S), Slope Length (L), Cropping Factors (C and P). My analysis is based on but not identical to the universal soil loss equation. My analysis does not attempt to produce a numerical value for the volume of soil lost, but instead used similar factors to produce a qualitative indication of erosion potential. In addition my analysis substitutes other data, such as soil permeability, agricultural use, or the absence of endangered vegetation for certain factors used in the universal soil loss equation. E = KA m S Figure 2. Stream power law. The second mathematical basis for my for my model is the stream power law (Figure 2). The stream power law assumes that the rate of erosion is proportional to slope (S) and the upstream catchment area drained through a location (A). This formula does not incorporate factors such as vegetation or soil, but offers a more realistic prediction of erosion due to fluvial processes (Lague 2014). Instead of using slope length (L), the stream power law uses the upstream catchment area (A), along with slope (S) to predict the erosion rate at a given location. My analysis substitutes upstream drainage area (A) for the traditional slope length (L) used in the universal soil loss equation. I believe this substitution offers a more accurate and robust representation erosion. Additionally, this method better reflects role drainage network shape has on spatial variation in erosion. My implementation of the stream power is based on the Fastscape landscape evolution algorithm. Fastscape is a powerful algorithm for modeling landscape evolution using the stream power law (Braun and Willett 2013). I imported elevation data from the 10m DEM of Maui and then used Fastscape to determine the slope and drainage area at each cell. I then used these results in a standard suitability like analysis to determine areas prone to erosion (Figure 3). n

4 4 Figure 3. Conceptual Model for erosion potential analysis. In addition to slope and drainage area my analysis incorporates the following: proximity to roads, soil permeability, agricultural land use, and the absence of endangered vegetation. This data was obtained from a variety of sources including the USDA as well as the Hawaiian state government and the County of Maui. I assume that all of these factors also contribute to erosion. Figure 4 summarizes the data sources and processing techniques used for my analysis.

5 5 1. Data Sources USGS 10m resolution DEM NASA 15m resolution Landsat Imagery USDA Soil Classification Data Maui County Road Centerlines. Agricultural Land Use Endangered Plant Abundance 2. Data Processing Elevation data from the DEM must be cleaned and formatted properly for use in the Fastscape model. The most significant processing is to convert the spatial elevation data from a 2d array to a 1d list. Fastscape takes spatial data using a one dimensional list rather than a two dimensional array to allow more efficient data processing. I have written several Python scripts in order to automate the process of converting two dimensional spatial into a one dimensional format. The procedure goes as follows: i. Determine width and height of study area (nx and ny) ii. Use python script to convert a nx by ny array into a nx*ny (nn) list of data iii. Use list nn and nx ny dimensions of study area to process data within fastscape model iv. Use nn and nx ny to reconstruct study area after analysis is complete Clip data to the extent of the Northwestern Maui study area using tools in ArcGis. Union data and assign qualitative suitability analysis variables based on susceptibility to erosion. 3. Spatial projection: WGS 84 Figure 4. Data acquisition and processing summary.

6 6 Results The results of my erosion potential model are presented in Figure 5, with areas of special interest shown in more detail in Figure 6 and Figure 7. In general the most intense sources of erosion appear to be concentrated near the coast, where development and agriculture are most intense. The interior, while steeper in general, is largely stable due to more extensive and diverse plant cover, permeable soils, and a relative lack of human disturbance. The most intense sources of erosion are concentrated in the wide belt of land between the heavily developed but largely stable shoreline and the undeveloped highlands. In this region a number of factors, both natural and manmade, combine to produce excess erosion. Unlike the mountainous highlands, these lower areas are relatively dry, which promotes dusty and impermeable soils that wash away easily during large storm events. Extensive agriculture in this region only exacerbates the problem, whether by striping the already unstable slopes of vegetation through grazing or by replacing more durable and diverse natural vegetation with crops. The most intense erosion is not associated with agriculture, but with development in this dry belt. Areas where roads and associated construction extend far inland, rather than being confined to the coast, appear to be the most intense, if relatively confined, sources of erosion. These areas of inland development would be an excellent topic of further study or field observation. Implementing soil loss prevention efforts here is likely the most realistic and promising strategy to curb erosion in the short term. Another, albeit more costly and environmentally problematic, short term solution is to build retention ponds to trap sediment at key points along the drainage network. This strategy is

7 7 an effective short term fix for reducing the sediment load on coral, but is untenable in the long run. Retention ponds can serve as a stopgap measure at best.

8 Figure 5. Erosion potential map based on suitability analysis. 8

9 9 Figure 6. Zoomed in view showing Lahina and KaNapali. Figure 7. Zoomed in view showing Kapalua and surrounding area.

10 10 Conclusions The results of my erosion potential modeling and analysis point to several potential options for mitigating soil loss and coral damage: construction of sediment retention ponds, restoration of native vegetation, and regulation of human development. In the short term, placing retention ponds, structures which are designed to temporarily retain water and trap sediment, at crucial points in the drainage network is the only feasible way to stop soil from reaching coral. In order to guarantee the long term health of coral, however, efforts should be made to restore native vegetation cover where possible and mandate land use practices to curb soil loss. Implementing these measures will take many years, but are the only way to strengthen hillsides indefinitely and guarantee the long term health of Maui s coral. Retention ponds offer a powerful yet expensive and labor intensive method to reduce sediment load. A retention pond is a manmade structure, similar to a dam or weir, designed to trap sediment during large precipitation events and remain dry under normal circumstances (Wolfe 2006). These structures are effective at trapping sediment, however they come with significant cost and environmental impact. Constructing retention ponds will likely require heavy machinery, the use of which may only further exacerbate the problem of soil loss. These ponds must also be emptied of sediment when they fill, meaning that roads to the such ponds are likely to see intermittent use for the entire life of the pond. While retention ponds may represent the only hope for drastically reducing soil loss in the short term, they are a brute force solution which cannot be maintained indefinitely. I propose using selective placement of retention ponds at critical junctures in the drainage network. If properly placed, a relatively small number of retention ponds could block a significant fraction of escaping sediment. In addition, the targeted nature of this approach allows for selective implementation in areas where coral is experiencing the most damage.

11 11 Retention ponds could protect coral in the immediate future and buy time for the development a long term solution involving broad but slow changes in land use. These changes in land use would have to involve both the restoration of native plants in relatively undeveloped upland areas and the management of human development near the shore. Vegetation protects hillslopes by both reinforcing soil with roots and by shielding soil from direct precipitation with overhanging leaves. Changes in land use caused by humans, especially the introduction of nonnative grazing animals, can lead to huge changes in vegetation cover and is often a major cause of soul loss. Grazing animals devastate vegetation and disturb soil, especially when no predators exist due to human activity or ecological isolation. Restoring sturdy, native vegetation to high grade hillslopes will help reinforce these hills and reduce sediment load to the ocean. Additionally, this vegetation had none of the long term maintenance requirements or environmental impact of retention ponds. Another aspect to consider in any plan for mitigating soil loss is human development. Resorts, commercial developments, and private homes are often designed without taking soil loss and erosion into account. Implementing reasonable guidelines on development, such as requiring large scale landscaping to include plants which are resistant to erosion, will also help to reduce soil loss and ensure the long term health of Maui s coral reefs.

12 12 References "Best Management Practices for Stormwater Runoff." (n.d.): n. pag. Laramie County Conservation District. Web< content/uploads/2011/04/ponds.pdf > Braun, Jean, and Sean D. Willett. "A very efficient O (n), implicit and parallel method to solve the stream power equation governing fluvial incision and landscape evolution." Geomorphology 180 (2013): Lague, Dimitri. "The stream power river incision model: evidence, theory and beyond." Earth Surface Processes and Landforms 39.1 (2014): El Swaify, Samir Aly, Edgar W. Dangler, and Clinton L. Armstrong. "Soil erosion by water in the tropics." (1982). Fletcher, Charles, et al. "Mapping shoreline change using digital orthophotogrammetry on Maui, Hawaii." Journal of Coastal Research (2003): Weesies, G. A., D. K. McCool, and D. C. Yoder. Predicting soil erosion by water: a guide to conservation planning with the Revised Universal Soil Loss Equation (RUSLE). Vol Washington: US Department of Agriculture, Agricultural Research Service, Wolfe, Donald L. "Hydrology Manual." (2006): n. pag. LA County Department of Public Works. Web< 0Hydrology%20Manual Divided.pdf >