Feature

Water Q&A: Opening the “black box” of modelling

SEI Senior Scientist Brian Joyce explains the value of water models in an interview, the first in a series from the SEI US Water Program.

Published on 20 March 2020

SEI Senior Scientist Brian Joyce explains a “serious game” at a modelling workshop.

Brian Joyce once made a vow: one marathon, every decade.

The SEI senior scientist has so far stuck to that quota – and also skis, gardens, and plays guitar “a lot”. But his real passion is opening up the “black box” of water models, by working with stakeholders to gather data, build future scenarios, and test out policies.

Joyce has worked with SEI’s flagship Water Evaluation And Planning tool – known as WEAP – to build more than 20 models that help policy-makers plan for an uncertain future under climate change.

His recent work includes building a model of Morocco’s main agricultural basin, working with Lesotho to analyse water security measures, and developing WEAP models for seven major river basins in Africa. All are built using SEI’s Robust Decision Support process, where Joyce and other SEI researchers work together with policymakers and stakeholders to create a shared mental model of available opportunities and potential trade-offs for various objectives.

Joyce talked about stakeholder buy-in, the value of models, and notable successes in a recent interview – the first in a series about researchers and their work in the Water and Ecosystems program at SEI’s US Center.

Learn more about the SEI US Water Program for Ecosystems and Livelihoods

Visit the SEI US Water Program

What was your first job?: My first job ever was painting houses. I started that when I was like 14 and then my friend, who was running the business, passed it on to me, so I ran a house painting business in college. I made good money over the course of a couple of months each summer – enough that I didn’t have to work during the school year.
How did you get into water modelling?: In college, for my undergraduate, I studied natural resources studies at the University of Massachusetts in Amherst. It was a pretty broad focus, as you can guess by the title. I remember talking with my advisor, who was sort of adamant that in natural resources, water was going to be the most critical one in the future. That resonated with me, but I didn’t know what to do about that.

I went into the Peace Corps shortly after that. I was in Senegal, West Africa and as I was there — working with people and getting a sense of what the challenges were — I encountered it directly that water is just a really big deal. If the rains don’t come, their crops don’t grow and it’s a big problem. They’re kind of living there on the edge.

From my advisor’s advice to me and from that experience, I decided that I wanted to get into water resources management, with a particular focus on agriculture. So then I applied to graduate programs and went to UC Davis.

I think models are really only useful if people can attach value to them. That value being that we can run all kinds of scenarios, asking questions of: What if we change this? and What if we did that? And we can see how the whole system responds.

Modelling can seem abstract. How do you get buy-in from stakeholders and connect it to real-world action?: It’s a really good question because I think we struggle a bit with that perception, of models being black boxes. It’s a personal mission of mine to demystify that.

I think models are really only useful if people can attach value to them. That value being that we can run all kinds of scenarios, asking questions of: What if we change this? and What if we did that? And we can see how the whole system responds. Then it becomes more about just gaining a sense of relative impacts, adjusting things here and there to see how a policy can benefit one sector or have an adverse impact someplace else. It’s less about the absolute numbers and more about gaining that intuitive sense of the impacts relative to one another.

Obviously you’ll reach people at different levels. There are technical people who get it right away, and then you have policy-makers who you communicate to on a level that makes sense to them.

Interested in the SEI US Water Program's work in Africa?

You can learn more about our water work in Africa here.

What are you working on right now?: Right now, I am working on a model of the Souss-Massa Basin in Morocco. We’re looking at the connections between water and energy and food production.

The basin there is very heavily agricultural. I think about 40% of the citrus that comes from Morocco is produced in the Souss-Massa. But it’s very dry, and they are pumping more groundwater than is being replenished. And as the groundwater levels decline, more energy is required to pump it.

It’s an important basin for agricultural production, so they’re building what will be the largest desalination plant in Africa that will support the city of Agadir in its first phase, and, in the second phase, will provide irrigation water for part of the basin.

We’re developing a WEAP model to look at the water system and the agricultural system, and we’re linking that to an energy model so that we can consider the full implication of different management strategies on water and energy and food. One thing that we will be considering are how changes in cropping patterns would affect water security. So perhaps they convert to crops that require less water.
Can you point a success you’ve had or something you were really proud of?: I guess I’ve got a couple of them. Domestically here in the US, we have been working in California with their Water Resources Control Board to use WEAP to evaluate different scenarios for setting flow standards as part of developing a new Bay-Delta Plan, which is a fairly massive undertaking in such a complicated system. It’s been really rewarding to see how our model can play a key role in this process.

In Lesotho, we worked quite intensely with the ministries of water and agriculture to train them in the use of a WEAP model that was developed for the country. Lesotho has an agreement with South Africa on how much water they deliver through the Lesotho Highlands Water Project.

Lesotho is pretty water insecure despite having a lot of water. It’s just not in the right place. We discovered that if they take just a little bit of that water that would otherwise go to South Africa, it could make a really big impact on water security in Lesotho, with relatively little impact on their deliveries to South Africa. So that was a big step forward I thought.

In terms of models directly influencing policy, more recently in Tanzania, we have been using a model of the Mara River basin to look at scenarios of water allocation in that basin. Tanzania is in the process of developing a water allocation plan that is informed by those scenarios.
Disputes over water are widespread and it can be a fraught topic. How do you handle that?: We take a neutral stance within that dynamic and we try to just provide an objective lens through which we can address these issues. And that requires us to work with all sides to develop trust in the tool we have created.

So going back to Lesotho: what had been happening there for years is that a tool that they were using to look at their water resources was developed by South Africa and it was very much a black box. So there was a lot of distrust on the part of Lesotho concerning the results that were coming out of this model.

After developing a new WEAP model with Lesotho, we shared the results with their South African counterparts. They reviewed them and agreed with what we were doing, so we were able to work with both South Africa and Lesotho to develop their trust in the model.
What’s next for you, in the world of water modelling?: We developed WEAP models for seven major river basins in Africa that cover about 40% of the land area. We were able to take one of them – for the Upper Orange River Basin in Lesotho – and apply it in a way that really drilled down and looked at some specific issues, like water security.

It’s been a really interesting process of starting large and becoming more and more focused in one particular area. I see that there’s an opportunity in other places, maybe using the other 6 models, to do something similar.
What do you most enjoy about your work?: I most enjoy when you can see that it clicks for people, when you know that people understand. I think that people struggle with the modelling for obvious reasons. It’s just not always really intuitive. When we work with people and you can see that people are starting to get it and understand the value of these tools, then it makes it worthwhile.

I spend a lot of time working on these models in isolation and when people see the value in it, I think, aha! I’m not doing it for nothing! It’s making that connection with people.
What do you do for fun outside of work?: I play guitar a lot. I don’t play in a band but I’m actively seeking one. I’m between bands, you could say. In the winter, I love to ski, both alpine and cross-country, and now it’s time to think about buying seeds for the garden. So I’m excited about that.

I made a vow that I was going to run at least one marathon each decade. I’ve done one in my 30s and one in my 40s and well, the next one is coming up, so I guess I got to start training.