Catherine Nakalembe: Enhancing Agricultural Productivity with Earth Observation
A conversation about increasing national governments’ knowledge and skills to monitor agriculture productivity with EO for informed responses to shortfalls in food production.
It is our pleasure to introduce Dr. Catherine Nakalembe, Assistant Research Professor at the University of Maryland. Dr. Nakalembe travels the world working with national ministries and regional agencies in East and Southern Africa to monitor agriculture with Earth observation (EO).
As Lead of the NASA Harvest Eastern Africa-Hub program and part of the NASA Harvest and SERVIR Global Applied Science Team, she conducts remote sensing training in the use of EO tools to assess and forecast crop conditions. Her EO capacity building portfolio includes the government ministries in Kenya, Rwanda, Tanzania, Uganda, and Mali, as well as regional agencies such as the Regional Centre For Mapping Resource For Development (RCMRD) and IGAD Climate Prediction and Applications Center.
Dr. Nakalembe started her career in Kampala, Uganda, where she earned a BS at Makerere University in Environmental Science. During this time, she worked on mapping encroachment in Mt. Elgon National Park. She went on to do her masters in Geography and Environmental Engineering at Johns Hopkins University and her Ph.D. Geographical Sciences at the University of Maryland. She wrote her doctoral thesis on “Agricultural Land Use, Drought Impacts, and Vulnerability: A Regional Case Study for Karamoja, Uganda.”
Dr. Nakalembe was the Program Assistant for the NASA Land Use Land Cover Change Program while working on her Ph.D. She also led the development of the first very high-resolution base map of two refugee settlements in Uganda using drone and satellite imagery, and GIS tools during this time.
In this Q&A, Dr. Nakalembe talks to us about increasing national governments’ knowledge and skills to monitor agriculture productivity with EO for informed responses to shortfalls in food production.
You are a pioneer in repurposing drones away from agricultural monitoring to the survey of refugee settlements in Uganda. What got you interested in technology and using it for social good?
I guess it is something we do every day with satellite imagery — we use the same data to study different things. To me, drone data were no exception.
I was primarily focusing on drought and agriculture, which are, and will continue to be, critical to monitor in refugee settlements in Uganda for the sustainable and equitable development of both the refugee and the host community. However, the most pressing needs for newer settlements are the accessibility to basic assets, such as clean water, shelter, protection, medicine, among others. A good baseline map is critical to study and monitor environmental changes; It can tell us where environmental features in or around settlements are, and how to optimize locations of basic assets was and remains critical.
Yet, baseline maps were missing for Oruchinga Refugee Settlement (Established in the 1960s) and BidiBidi Refugee Settlement, which had just been established (2017).
I was in Uganda for fieldwork to assess the 2016 drought in the North East when I was asked by the Disaster Risk Advisor to develop ideas around data and mapping. And I proposed Drone Mapping since at the time we had two drones we were using primarily for agricultural research. So, I have to say 2017 must have been the most productive year in my life, besides all the training and fieldwork in Bidibidi, I also had my Twins that September.
Both refugee camps faced tremendous challenges. At the time, Oruchinga’s 6,700+ population were plagued by overcrowded schools, and conflict over scarce land and resources. The BidiBidi Refugee Settlement opened July 2017 had a population of over 280,000 by December 2017. Access to land for agriculture and clean water are some of the problems the communities face especially in which such a rapid change in the human population in a region that is already environmentally and social economically strained. Using field-collected drones and satellite imagery seemed like the best option to get both refugee settlements mapped quickly and efficiently. I subsequently developed the first-ever very high-resolution base map of the Oruchinga and BidiBidi Refugee Settlements.
As partnerships are important, I brokered a stronger link between Makerere University’s department of Geo-informatics and Uganda’s National Emergency Coordination Center with the University of Maryland and the United Nations Development Programme as supporting mechanisms. I also trained local University students (From Makerere and Islamic University) on the collection and use of drones and satellite imagery for mapping infrastructure for development planning in refugee settlements in Uganda.
The mapping work I led inspired others to do the same. Today, most, if not all refugee settlements, have been mapped in Uganda.
You focus a lot on EO capacity building for staff members at national agricultural ministries in East and Southern Africa. How successful are these training workshops in terms of generating insights by local ministry staff or informing policies?
I would say highly successful! My primary focus has been to get as many government agencies that focus on agriculture and food security monitoring to access operational, reliable and high-quality data and systems that can allow them to work independently. I believe that this approach significantly helps to fill the data gaps in many countries that can’t be filled otherwise; and once the team(s) are on board, the appreciation of EO data is evident. Training government officials are more sustainable in the long-run as trainees can incorporate outputs in their workflow and it becoming part of the institutional culture.
These quotes from my training events speak volume:
“This is exactly what we’ve been looking for.”
“Why didn’t you start with us?”
“When can we have another training?”
“I have so many questions for you.”
“We need this.”
“Wait till you see our next report!”
The training translates into the regular use of satellite data to assess crop conditions and forecast for agricultural decision-making. Some countries have implemented signature reports that are based on these assessments for the ministry department. In Uganda, for instance, reports such as the “Disaster Risk Financing Report” are now regularly used to brief Government Ministers and inspire more partners to develop a concerted response in case of emerging problems.
. . . What are some lessons learned for increasing the national government’s knowledge and skills?
Trust is of utmost importance; Building trust can help one overcome many challenges. Once the teams feel they can work with you and rely on you to support them, then knowledge transfer is easy. Sometimes a quick reply to questions posed on WhatsApp is all that’s needed to remove a technology barrier.
Patience is needed. Remote sensing and GIS are two subjects that officers do not have to use frequently on the job, but I often have to spend a lot of time working in, and teaching, basic programs like Word or PowerPoint to improve participant skills of drafting reports. It can take a really long time to complete what one might consider a “simple task,” e.g., “copy + paste.” It’s only until everyone is confident that they can write and complete their own reports that you will have succeeded.
If you speak to someone from the Global South, you’ll often hear that EO data are hard to come by. Yet, EO data are freely and readily available via organizations such as ESA and NASA. What do you think needs to be done to raise further awareness?
More applied science projects and programs can help with this. The data become more accessible when we work with the intended end-users.
We use NASA data — which are open data — for all our EO training on operational systems to ensure that the data remain accessible for developing nations to use for managing issues such as natural disasters, agriculture, and natural ecosystems. Our projects help to bring awareness to open knowledge for practical applications in the Global South.
Based on your new NASA SERVIR project, semi-automated baseline datasets are necessary for more accurate agriculture monitoring. What are the most significant issues you face in collecting these ground-referenced data?
The cost to sustain and continue the data collection is an issue. The methods, the tools and the guidelines to do this correctly at a reasonable cost exist. For instance, one can work with existing structures (extension agents) to offset most of the cost and use the guidelines developed by GEOGLAM for Joint Experiments on Crop Assessment and Monitoring (JECAM). However, if that data continues to be collected on a project to project basis, then we lose the continuity and consistency of the data. This is an obstacle, especially if you want to monitor change.
What is your hope for the Earth observation for National Agricultural Monitoring Project? If successful, what impact will it have on food security in East and Southern Africa?
A successful project will mean better reference datasets through improved and systematic field data collection that can support on-going and future work for mapping crop types/areas and yield in smallholder agriculture. Better datasets mean informed and improved decisions that can save lives and improve livelihoods through food security. I hope to develop scalable methods and workflows that can be used in other countries on the African Continent.
Geospatial technology and Earth observations are essential for solving global challenges and creating a sustainable future. However, data are not always within reach of countries, especially in the global South. In the context of lessons learned and doing things better, what is needed to lessen the data gap?
Education and training. If more scientists, researchers, and businesses from the Global South are comfortable and capable of working with, manipulating and harnessing the value of EO, the more likely that the full potential of EO will be realized. There are so many problems we can apply EO to and most of these can only be perceived and correctly studied/addressed with local knowledge.
It’s also not only about the data existing. The data have to be in the right hands and skills. We can truly address these problems with knowledge and training.
