Space in Africa had a chat with Ufuoma Ovienmhada on this and how the use of Earth observation satellite data has helped to curb the effects of invasive species, amongst other things.
Ufuoma Ovienmhada is a research assistant and a second-year graduate student at the Space Enabled Research Group of MIT Media Lab. She graduated from Stanford University in 2018 with a Bachelors in Mechanical Engineering. Ufuoma has worked in the IRIS lab for sustainable design and at Public Lab, an organisation that develops community science tools and practices for environmental exploration. She has also worked at an Ivorian company called Somated, as part of the Stanford Seed program to support entrepreneurship in emerging economies. Also interested in education, Ufuoma has worked with several STEM initiatives for underrepresented groups—most notably StreetCode Academy of Palo Alto as a Lead Curriculum Facilitator. With Space Enabled, Ufuoma researches opportunities for space technology to support healthy and equitable socio-economic and environmental ecosystems in West Africa.
So, how long have you been with Space Enabled?
A year and one month, I’m a second-year Masters student right now. I am a part of the first cohort of graduate students of Space Enabled.
Is it just a Masters’ degree programme or it extends PhD?
Uhm. It could be both. So, the Space Enabled Research Group is part of the MIT Media Lab, which offers both Masters and PhD Degrees. The Masters is a two-year degree with the thesis, but then you can apply for a PhD degree as well.
So, will you be doing a PhD Degree?
Potentially. (laughs) I’m interested.
Okay, is the group still open for applications?
Space Enabled? Each faculty has the disposition to open or close their search each year. People can refer to this website to check if applications are open for Space Enabled or other Media lab research groups.
Tell me about your research.
I study the application of Earth Observation Technology for sustainable environmental management. Specifically, for my Master’s Thesis, I am looking at the management of invasive plant species in Benin. What are invasive species? Invasive species refer to plants or animals introduced into environments they were not meant to grow. Sometimes, this isn’t a big deal, but sometimes these invasive species are more competitive than local species, so they can kill them, destroy habitats and cause massive economic repercussions. The plant I am studying right now is the water hyacinth. It’s a plant that’s native to the Amazon Basin, in Latin America, but was introduced into West Africa and Africa in general about a few centuries ago, presumably, during colonisation, and the plant has wreaked havoc in many countries all around the world. Right now, we’re looking at it in Benin where…
Benin Republic, or Benin in Nigeria?
The Benin Republic, but they have the same issue in Nigeria as well. The plant has a lot of devastating consequences in the ecosystem: it creates a breeding ground for mosquitoes that may carry disease; it also grows really densely, so the number of water hyacinths can double in surface areas in just two weeks, clogging the entire river, which makes it very difficult to cross. The weed gets caught in the boat motor, so you can’t really move through the lake, because it’s just so thick. Also, the plant absorbs lots of oxygen when it decomposes, which kills fish or any other mammal that needs oxygen to survive. It’s a huge issue. In one year, it caused damages of USD 84 million for the local community.
In the Benin Republic alone?
Yes, in one year. That is an estimate though, but yes.
Wow, so how are you using Earth Observation to tackle that?
So, the Space Enabled research group of MIT, led by Professor Danielle Wood, started a collaboration with Green Keeper Africa a bit over a year ago. Green Keeper Africa is a company based in Cotonou, in Benin, led by director Dr Fohla Mouftaou. They pay local community members to harvest the plant and transform it into a fibre that’s good in absorbing oil-based pollution. It is something they have been doing for a few years, which is awesome because now they are taking this plant that was a nuisance and turning it into something positive.
Can you throw more light on how you are using satellite data to deal with this problem?
At first, we were using the Landsat series. Landsat is an optical data that dates back to the ’70s in some regions, and using that, combined with graphical indicators such as NDVI, we can identify the presence of vegetation over the water which gives us an approximation of any vegetation, but that’s not specific to the water hyacinth. To get the water hyacinth is more complicated. Other people would use things like hyperspectral imagery, but that’s expensive, and the free versions usually have low resolution. So I developed, or rather, applied an existing technique called “Change Detection” in a new way to distinguish the water hyacinth from other species. Using this method, we’ve been able to estimate the surface area coverage of water hyacinth from 1985 to present, for each month using satellite data.
You mentioned Landsat, and it gives you like 30 metres resolution, right?
It is difficult trying to study a specific plant with Landsat. I was expecting you to mention better resolution imagery, like 0.5?
We use Sentinel too, which has a resolution of 10 meters, but that’s only been available for the past four years. The reason why we can use Landsat effectively here, despite the weaker resolution, is that the water hyacinth grows in these clusters that are often up to, if not greater than 30 metres in size. Also, the water hyacinth is a floating plant, you can see it in one location in an image, and a few days later, it’s moved a few metres. So, these are pretty distinguishable characteristics of the water hyacinth which makes it readily detectable by most resolutions.
So, this particular plant could grow or go with weeds, right? Maybe other plants, how do you differentiate them?
Sure, that’s a good question. So, this is where the Change Detection technique I mentioned earlier comes into play. The water hyacinth has high sensitivity to salinity because it is not native to the region. You don’t see the same type of growth/death cycles in other water plants on the Lake. So, on Lake Nokoue, we see a peak period of the water hyacinth growth during the rainy season and one period, during the dry season, where it completely dies. Other plants that are native to this region do not see that trend. Taking advantage of this, we calculate the NDVI anomaly over that peak/low period and calculate the NDVI change over time, with respect to a reference period. In geographic areas where I see an average decrease in NDVI, I make the approximation that it’s probably water hyacinth, but in other regions where we see a relatively constant NDVI over this time period, I hypothesize that it is local vegetation.
Have you been able to validate it with ground data?
That’s where the drone imagery comes in. The drone we are using has a much higher resolution, at 3 centimetres, than the satellite data so we can use this imagery to distinguish between water hyacinth and other vegetation and validate the approximations used from the satellite data. We practised flying drones in August and are currently planning more drone flights early next year to complete this step.
So you are hoping both will tally?
I don’t think I understand this question
Yeah, I’m hoping it does (laughs).
That’s interesting. Why did you choose Benin?
We didn’t choose Benin …
Is it because of the local partners?
Yeah, definitely. The Space Enabled Research Group has a philosophy that we shouldn’t go places we’re not invited. So, it’s very important to have local guidance and local leadership on things like this, you know, in order to make it sustainable. This project began because of the connection with Dr Fohla Mouftaou
Who is funding the project?
So, a combination of MIT Media Lab, a grant from NASA, and miscellaneous grants from other MIT sources.
Okay. what’s the timeline for the project?
Good question. From my Masters’ thesis, I’m hoping to prove the validation of this satellite technique for identifying the water hyacinth and the design of a low-cost sensor kit. At the same time, we’re working to create an online model that can predict where the water hyacinth will be. We’re working with the company, Blue Raster -that specialise in creating user-friendly web interfaces. I foresee by the end of my Masters’ thesis, which is June 2020, having a pretty decent prototype that integrates both the satellite imagery, drone imagery, and this water quality sensing as well.
Are you the only one working on this project or is it a lab thing?
I’m the lead graduate student on it but the project has been made possible with lots of support from NASA professionals and undergraduate researchers.
Thank you for your time, Ufuoma.