Progress Report

Article Chosen: Modelling future landscape change on coastal floodplains using rule-based GIS by Iain Brown

Summary: Flooding, especially along the coastline, is an important issue today due to the the changing climate. Brown studied the effect of flooding on the coastal plain using elevation raster data analysis. Since sea levels are rising, it is important for communities to adapt to the changes by mitigating the effects with policies, such as sea walls or changing land use. Brown's study analyzes four possibilities: maintain existing line, small-scale realignment, large-scale inland realignment, and no defences. Elevation data layers were used to see if the water will go over a point of high elevation, thus showing if the sea level policy strategy was effective. When different types of land are flooded, the land type changes because of the increased water, which affects the effectiveness of flooding buffering provided by the land cover.

Policy 1: (maintain existing)

Keep current policy
Add a few defense features where absolutely necessary
Focus on protecting settlements

Policy 2: (small-scale)

Minor inland adjustments
Focus on protecting habitats

Policy 3: (large-scale)

Move the defense line further inland
Return to a more dynamic natural coast
Minimize the need for man-made defenses

Policy 4: (no defense)

No defenses at all
Acts as a control

Brown's Methodology

Our Support/Critique

Step 1:

Getting the Data

Brown collected six data layers (sea level, elevation, land cover, tidal and flood levels, policy options, and buildings) to complete this paper. Since Brown is a UK academic and has financial support, he is able to gather all of these datasets.

Because of our financial and geographical limitations, we were only able to gather a few of these datasets in full, which creates challenges in reproducibility.

Step 2:

Elevation Data

Brown created an elevation layer that only had the topology of the region and sea-walls, removing things like vegetation and buildings.

This could be a potentially negative decision because buildings and vegetation have effects on how the water flows. Buildings could act as sea-walls, making them an important part of the landscape to consider. It also would be important to know which buildings flooded after an inundation event, which would be difficult to analyze if the buildings were removed in the first place.

Step 3:

Hydrological Analysis

Brown uses a modified version of watershed analysis, essentially, running the process in reverse to model flooding. Brown first used the "fill" tool to fill any gaps in the raster layer. The pour point is the point of high elevation that acts as a decision maker between flooding and not flooding, as seen in the diagram below. If the water reaches or exceeds the elevation of the pour point, then the land behind the pour point will flood.

This process is very similar to the watershed delineation project we did for Special Works Project 2. Watershed delineation is a straightforward process that can easily be replicated. Brown provides a creative use for delineation by using the tool to analyze flooding, which is the opposite of water flowing down a watershed.

Step 4:

Adding Barriers

To mitigate against flooding, sea walls and other forms of barriers are used as defense mechanisms. Brown created raster layers that included sea walls and barriers. Since the sea walls have higher elevation, they act as pour points that prevent sinks further inland from flooding. Brown used a paid company to manipulate the raster layers to reflect the different policy options. This process was repeated for all four policies.

The process of manually changing the raster elevation layers is possible in ArcPro, but has to be done cell by cell. Because this is a small study area, this is possible, but not recommended for large areas. It is a good idea to artificially add the sea walls into the raster layer for modeling purposes because it is costly to build sea walls, especially when it is unclear if they will be effective. Our only hesitation is that changed elevation will not react in the same way that an actual sea wall would, but is the best option for a model like this.

Step 5:

Transition Rules

To make this study useful in the future, Brown used transition rules to "predict the future state for each grid cell" based on the current land-cover and the tidal/flood levels. This looks at how the land cover would chance based on the level of flooding. These rules also included anthropogenic changes. The transition rules are shown in the table below.

These transition rules make this study more useful for future studies. Especially given the environmental value of salt marsh, it is important to see how the land will change as a result of rising sea levels due to climate change.

Step 6:

Results

Validation: To make sure this methodology was correct, Brown compared a zero centimeter sea level rise with no policy change to the original data set. Because the results of these were the same, Brown knew his methodology worked.

Future Landscape Change: This model was run for each of the policies for low and high flooding values for a total of eight tests. The graphic below shows the resulting land cover under high sea-level rise conditions for each of the policies. Because of the importance of salt marsh as a buffer against flooding, Brown is particularly concerned with how the policies affect the level of salt marsh and the environmental costs to other types of land.

"A general implication is that the more radical the policy option is compared to Policy 1, then the greater the potential for landscape change, but this may also bring greater flexibility in adapting to higher rates of sea-level rise."

Brown's analysis seems to ignore the effects of Policy 2. Policy 1 and 4 do not help reduce flooding and Policy 3 is the most cost effective. However, when looking at the number of buildings flooded, Policy 1 and Policy 2 have the least amount of flooded buildings by a significant amount. This leads us to question what is important to consider: the amount of money spent or the effectiveness of the policy?

Step 7:

Cost Analysis

Defense costs are calculated by length of defence, maintence, and capital costs to build and mainain the defense structures. Policy 3 is the most cost effective because it requires only 1km of new defense structure and is effective at preventing flooding.

Cost analysis is an important step in any scientific study. It makes sense that Policy 3 is the most cost effective because of the limited amount of barrier building the government would have to do. Policy 3 also takes into account both settlements and environmental health, making it externally beneficial. Health of the environment should be taken into account in cost analysis procedures.

Conclusion

Brown's methodology is reproducible, but we have two main critiques of Brown's work:

Buildings and vegetation are important to consider
There was no policy chosen as the best policy

Our Progress and Projected Timeline

We are currently on Step 2 of Brown's methodology. There are limitations on how much we can edit the raster layers because we don't have access to the algorithms the author used (ie TerrainFit Algorithm).

At the rate we are going, we would need at least two more months to complete this project. We have just gotten (most of) the data to a usable place to begin the geoprocess. Once we begin reproducing the geoprocess, there probably will be more obstacles that take time to resolve.

Page updated

Report abuse