Mapping T&T

St Helena Trinidad, Flood Mitigation Option #1: Levees!

In the previous post I showed that the main reason for flooding in St Helena is its location in the low-lying terrain between the Caroni and Guayamare Rivers. This post will explore one option of flood mitigation in the area – levees. Levees have been used for centuries to protect areas against flooding. Portions of the Caroni River banks have already been raised to protect nearby villages, such as at Ibis Gardens. However these are only able to provide protection against lower frequency storms.

In the 2D hydraulic model of the Caroni River basin, I raised the south bank of the Caroni River from the Uriah Butler highway to an agricultural road upstream of St Helena. I ensured that the top of the levee was higher than the hypothetical 100-year water surface elevation along the entire reach. The map below shows the proposed alignment.

I ran the 100-yr storm on the model and compared the results as shown below.

Hypothetical 100-year floodplain at St Helena, Trinidad

Hypothetical 100-year floodplain at St Helena, Trinidad with potential levee, shown by the distinct outline of the south side of the Caroni River floodplain

As shown, there is much less flooding in St Helena (and Kelly Village) with the levee in place. There is still some flooding along the Guayamare River and through the low-lying areas of St Helena, but the difference is obvious. This does not mean that levees are the answer.

Levees are expensive to construct and maintain. Levee overtopping can be catastrophic to areas they are supposed to be protecting. Additionally, the construction of a levee inevitably increases flooding upstream as well as to the north of the Caroni River (see the western tip of runway at Piarco airport).

Any flood mitigation project should have No Adverse Impact on other areas. In future posts I will explore other options for decreasing the flood risk in this area.

St Helena Trinidad, Flooding Causes

In a previous post I showed a hypothetical floodplain map of St Helena, Trinidad.

Hypothetical 1% Floodplain Map of St Helena, Trinidad

This is a lot of flooding! St Helena is one of those areas that are notorious for flooding whenever rain starts to fall in the area. But why? The answer is mainly its location.

St Helena lies between two rivers: Caroni River to the north, and the Guayamare River to the south. That means whenever levels are rising in either of these rivers, there is potential for flooding.

Secondly and more importantly, St Helena simply lies in a floodplain. In the map of Caroni River Basin Topography at St Helena, the village clearly lies in the historic floodplain of the Caroni River. It is not on higher ground and so is significantly more susceptible to flooding.

Here is a closer look at the St Helena area. There is a very large low-lying area between the Caroni River and Guayamare River that will be the first to be inundated. The elevation along the dotted line shows that once the banks of the two rivers are overtopped, the area in the middle will quickly be flooded.

Lastly, where does the water come from? The St Helena area is mainly flooded by the overtopping of the Caroni River (and Guayamare River) banks. The hydraulic model of the Caroni River basin shows that floodwaters leave the Caroni River, and choose the path of least resistance, through the low-lying area of St Helena.

So what can be done to save St Helena from future flooding, apart from taking the entire village and moving it to higher ground? I will try to look at some alternatives in future posts, using the hydraulic model of the area.

Caroni River Basin 2D HEC-RAS Model Overview

I had a request to see the model that I used to create the floodplain map for St Helena in my last post. I used RAS 6.6 to develop the model along with all the inputs I described in my previous posts: Terrain, Land Cover, and Soils. Here is a brief overview of the model.

2D Mesh

The 2D mesh is what ties all of the inputs together. I have an older post about it here. A 2D mesh is necessary to model the Caroni River as the floodplain is very wide and flat, with multiple flow paths and situations that cannot be handled by a 1D model. The objective in building the mesh is ensuring that all hydraulically significant features are “captured” by the cell faces. In other words, any topographic feature that affects the flow of water needs a line on top of it. Road embankments are typical examples, as well as dams or levees. Streamlines are also important, which are derived from the terrain.

Terrain

The Terrain informs the model which way the water should flow. Making connections through obstructions in the LiDAR surface is important if structures exist that convey water.

Roughness and Infiltration

The Land Cover layer is used to apply Manning’s Roughness values to the mesh, which influence how fast water travels over the surface. Fast surfaces such as roads or parking lots have a low value, while slow surfaces such as forests have high values. For Infiltration I used the Curve Number method because of its simplicity and widespread acceptance. Infiltration dictates how much water seeps into the ground and does not contribute to runoff.

Boundary Conditions

As I mentioned in my last post, I made up a rainfall depth as rainfall data is unfortunately not readily accessible. Ideally statistical analyses can be performed on the rainfall gauges in the watershed to apply the most appropriate depths varied spatially. The outlet is set to the approximate high-tide elevation, based on a tidal gauge near the mouth of the Caroni River, which provides the most conservative floodplain boundaries.

And that’s the quick and dirty rundown. Happy to expand on any of these topics. There are several things that can be done to improve the model. All of them are linked to making data free and accessible for all, which will go a long way in fostering a spirit of innovation and development that can only benefit the country.

St Helena – Trinidad Floodplain Map

Let’s talk about floodplain maps. Floodplain maps have many uses, including guiding development and zoning, identifying the most vulnerable areas for emergency preparedness and response, analyzing property risk, and water resources management. Here is a hypothetical floodplain map of St Helena, Trinidad.

If you have been following along with all of my mapping posts so far, you may have noticed a hydrological theme. Adding some hypothetical rainfall to the terrain, land cover, and soils of the Caroni River basin can give us some hypothetical floodplain boundaries using a hydraulic model. I say hypothetical because rainfall data is not freely accessible to the public, which is very unfortunate and stifles innovation and development.

Anyway as you can see there is a lot of hypothetical flooding. It does seem that agricultural fields are taking most of the damage. I chose the Tabloid paper size with a 1:10,000 scale to show the area in more detail. Adding roads and their labels, along with aerial imagery will hopefully help the user easily find their area of interest. An issue is that a lot of road names are missing or may be incorrect. I took some layout ideas from the FEMA Flood Risk maps, although these maps tend to be much bigger and use a smaller scale for built-up areas like this.

The symbology of the floodplain was a challenge. Water should obviously be blue, but that limits the amount of classes that can be used before its too difficult to distinguish. It would have been simpler to just add the floodplain boundary as a polygon, but I wanted to add more detail. I chose depth breakpoints that seemed reasonable to me but more thought is probably warranted here. Other parameters such as Water Surface Elevations, Velocity, and Shear Stress can be exported from the hydraulic model.

How can this map be improved? What can be done to alleviate flooding in this area? What other information would be useful to the residents/ community leaders/ government officials to to reduce the flooding risk and loss to property/ life?

Maracas Watershed 3D Map – Take 2

I wanted to revisit the 3D map of the Maracas watershed I did a few weeks ago, but this time using the Shaded Relief symbology that I experimented with. This map uses the default hillshade parameters. It wasn’t necessary to accentuate the vertical features as this is a more localised area.

This map shows the watershed in more detail than previous, as the hillshade makes it easier to visualize the changes in elevation. Also took the opportunity to upgrade the font.

Hillshade Testing

When I was putting together the map showing the Topography of the Caroni River Basin, I go up to speed on 3 important parameters when generating a Hillshade: Z Factor, Azimuth, and Angle. Here is a quick comparison of these factors.

The Z factor is used to exaggerate the vertical features of the terrain by multiplying the elevation values by the factor. The first image shows the comparison of different Z factors on the Caroni River basin terrain. The Z Factor of 1 shows the original values as they exist in the dataset. The changes in elevation are difficult to see, especially in the southern part of the watershed where there is less variation. As the Z factor is increased, these features become more prominent, but at the expense of losing detail in the Northern range as the hills begin to block the light source.

Z factor comparison on the Caroni River Basin Terrain — Z Factor Comparison

Azimuth refers to the sun’s position with respect to the subject. An azimuth of 0° means the sun is located to the north, 90° is to the east, 180° is to the south, and 270° is to the west. The 0° azimuth creates a lot of shadow south of the Northern Range, while the 180° illuminates it too much. The default azimuth is 315°, which is a north-west position of the sun.

Azimuth comparison on the Caroni River Basin Terrain — Azimuth Comparison

Angle (or altitude) is the position of the sun above the Earth’s surface. A 0° angle means that the sun is on the horizon, so you can think of sunrise or sunset, when not much land is being illuminated. 90° means that the sun is directly overhead, which results in no shadows. The default angle is 45°, which is a nice balance of light and shadow.

Angle comparison on the Caroni River Basin Terrain — Angle Comparison

The final map of the Caroni River Basin Topography uses the default azimuth of 315° and angle of 45°, while the Z factor is set to 5. It is worthwhile testing these parameters to see which combination best highlights the features you want your audience to notice.

Caroni Watershed Topography

A map showing the topography of the Caroni River Basin. Topography is the study of the land surface, which includes its physical features such as hills and valleys.

This one isn’t too bad and I learnt a few things about hillshade which I will hopefully share in a future post. The purpose of this map is to show the differences in elevation across the Caroni River basin. I used the default Elevation colour ramp. Its a pretty big improvement over my first draft below.

I switched up the basemap to the World Topographic map to go along with the theme. The World Hillshade layer is also included which gives context to the area outside of the watershed.

I stuck with the serif font since it looks more polished. I didn’t want to place any labels over the dataset so hopefully the lines make sense. I thought that users would want to know which valley in the Northern Range is which since they are so prominent. I also added the labels to the surface features that may stick out.

It is interesting how much higher the Northern Range is than the Central. The Caroni River is fairly obvious flowing down the plains which appear incredibly flat, and if you squint you can maybe see the Caroni-Arena Dam and Piarco Airport.

Another thing that I learnt was the advanced labeling on the Legend. See my draft below for how it used to look. Now I know how to add rough intervals to the colour ramp. My only complaint is that is didn’t seem to want to use the Elevation colour ramp so the colours are slightly off.

Draft Caroni Watershed Topography Map — My First Draft…Not as informative but very colourful!

Caroni River Basin Soil Map

A Soil Map of the Caroni River Basin in Trinidad

The purpose of this map is to show the soil textures in the Caroni River Basin. The dataset source is here: https://www.arcgis.com/home/item.html?id=1b97fe85db144d19ae71b60fa1bbf6ef which appears to be linked to this study: https://sta.uwi.edu/ffa/sites/default/files/ffa/USDA%20soil%20Taxonomy%20Upgrade-Trinidad%20and%20Tobago.pdf

Similar to the Land Cover Map, I had to dumb down the data a lot to get to this stage. Soils have a billion different properties that distinguish them. As a water resources engineer, I mainly care about how “hydraulically conductive” they are, meaning how well water can pass through them. Soil textures play a big role in how conductive they are, so that is what I have chosen to present. You can imagine in the case of water falling on sand, most of it will infiltrate into the ground, whereas water that falls on clay will mostly stay on top of it.

I tried to be clever with the colours here. Sand is usually shown as yellow so that is what I chose for sand. I chose red for clay since it is not good for infiltration and sometimes clay is red? Then the remaining primary colour is blue so that went to loam. Blue and yellow make green so that is sandy-loam, and blue and red make purple so that is loamy clay. The goal is for the user to easily distinguish the colours and possibly even guess what they mean.

Clay makes up the majority of the soils in the watershed followed by sand. It is interesting that the Northern and Central Ranges contain the clay, while the Caroni plains are mainly comprised of sand. This makes sense as runoff from the ranges is deposited in the plains which makes it good for agriculture. The strip of loam along the east-west corridor also stands out. Possibly has something to do with the development?

Had a hard time deciding where to put the map title. I think it ended up in the right spot.

Caroni River Basin Watersheds

A closer look at the watersheds within the Caroni River Basin

This is just a zoomed in version of the full Watersheds map https://trinimapping.home.blog/2025/01/11/trinidad-watershed-map/. The full dataset can be viewed here: https://codepen.io/leelum1/full/yQbBBr. The purpose of this map is to show the Caroni River basin in greater detail, along with the main rivers in each watershed.

Technically the Guayamare and Cunupia do not flow into the Caroni River, but I’ve included them as they do drain into the Swamp. Similarly, only half of the Port-of-Spain watershed flows into the River. The biggest watershed contributing to the Caroni River is the Cumuto watershed, which originates in the Central Range. The smallest is the Tunapuna watershed. The Caroni-Arena Reservoir watershed is the entire area that drains into the reservoir.

I experimented with the fonts again. Not really sure that I like how this looks. I was also very tempted to change Gulf of Paria to Gulf of Trinidad in honour of current world events.

Trinidad Land Cover

A Land Cover map of Trinidad

This dataset is sourced from: https://data.fs.usda.gov/geodata/rastergateway/caribbean/index.php. Obviously the major land cover is trees. I simplified the dataset as I didn’t think I need to post every single type of tree in the forest. The original dataset is pretty detailed, and free! (Which is no surprise as its from a US source). I would be interesting to see how this compares to a land cover map developed by the local government.

I found this map on the internet https://www.researchgate.net/figure/Land-use-cover-map-using-of-Trinidad-developed-using-the-Maximum-Likelihood-classifier_fig1_254306325 which looks pretty similar, although with pretty harsh colours. They also most likely classified the old sugar cane fields as Cultivated whereas I set them as Herbaceous.

This map https://www.researchgate.net/figure/Trinidad-land-use-based-on-classification-by-Helmer-et-al-2012-illustrating-location_fig1_336873618 seems to present a lot of information but it is pretty unuseable as nobody can distinguish that many colours. There’s a couple other land cover maps out there that I also should not be criticizing. What we need is a map/data source from an official entity.

I tried experimenting with the fonts. The one thing I remember from my RHET105 course at UIUC is that serif fonts are good for digital display, sans-serif good for print. I don’t know if that’s still true or even I imagined that, but here is a map using a serif font. I think it looks pretty sharp, especially the labels of the surrounding water bodies. I also got pretty good at spinning my pen around my fingers in that course.

Added the pie chart directly in ArcGIS. Learnt that pie charts are not available for rasters, so had to convert the dataset to a vector layer to have the option.

I also added some annotation to the areas that might stand out to provide more context such as the wetlands and reservoirs. I like how the information box came out. Spoiler: its just a bunch of white boxes grouped together.