Perspectives on CIMSS from Director Tristan L’Ecuyer

April 22, 2019 | Jean Phillips

The Cooperative Institute for Meteorological Satellite Studies (CIMSS) Board of Directors met in Dec. 2018 and appointed Tristan L’Ecuyer as the institute’s next director.  He began his role in January 2019.  L’Ecuyer is well situated to lead CIMSS into the future, bringing 20 years of experience at the “intersection of satellite remote sensing and climate science” to the position.  Our Q&A with him follows.

Tristan L’Ecuyer, Director of the UW-Madison Cooperative Institute for Satellite Meteorological Studies. Credit: Eric Verbeten

You are just over 100 days into your new role as Director of CIMSS. What have you learned about CIMSS that surprised you or that perhaps you did not know? Are you where you thought you would be at this point in your transition to the leadership role?

The short answer is that things are going better than I had imagined. One of the things that I have learned is that I did not have an appreciation for how cutting edge the research really is in many areas, even though I knew in the back of my mind that there was a wealth of great research and science being conducted at CIMSS.

Over the last 100 days, I have been meeting with as many people as I can. I have learned, for example, that some of the data products and the interfaces between the algorithm developers and the end- users, like the National Weather Service, are far more advanced than I had imagined. I was aware of those activities, but as I have learned more detail, I am even more impressed with the people in CIMSS and the programs they have developed.

I had high expectations to begin with, but I would say that the reality has exceeded my expectations.

Talk about your vision for CIMSS.

Again, there is a short answer and that is I feel more confident than ever that my vision for CIMSS is a good fit.

Initially, my vision was based on my previous knowledge, obviously wanting to continue, and embrace, the strengths of the institute in calibration and validation, dataset production, supporting end users, state-of-the-art satellite data visualization, and disseminating data to the community. I also had the sense that we will not survive if we only do the same things, in the same way, as we have done them in the past. The world of satellite meteorology is evolving. The world of data products and their uses is evolving. We need to lead that evolution.

As part of that process, I was hoping we would embrace new computational techniques for observations that we already have and develop new applications such as data assimilation and more advanced warnings for severe weather, to name a few, thereby expanding that portfolio. And explore new climate applications for our data products or applications in the insurance and risk management industries.

These are all things that I had hoped we would be able to do and in speaking with many of the investigators within CIMSS, I have learned that people are further along in that vision than I expected. We are already beginning to lead the transition to the next generation of satellite products and data uses that I believe will serve to grow the institute as opposed to worrying about its decline.

Another aspect of my vision was that in order to expand the user base of our data products we should  build more effective bridges to other agencies. CIMSS is primarily supported by the National Oceanic and Atmospheric Administration (NOAA), but there are a number of investigators who have funding from other agencies such as NASA or the Department of Energy. One of my goals is to make sure that the satellite products we generate are also of maximum use for these other agencies.

As an example, the recent National Academy of Sciences (NAS) Decadal Survey for Earth Science and Applications from Space includes a plan to move away from really large, expensive satellites to smaller platforms that observe individual atmospheric parameters. In order for that information to be useful, you need to add context – you need to know what else is going on in the atmosphere as those measurements are gathered. This is where the geostationary satellites come into play, especially our newer satellites (that CIMSS has supported since the very first one was launched). I feel that we have an opportunity to use the products we have developed to provide support for the new NASA missions in the next decade and beyond.

So, coming back to the point – what have I learned – these activities are already happening. Our cloud research teams, for example, are already generating subsets of their datasets to support other missions. Others are exploring advanced computational techniques for detecting severe weather and more rapidly identifying fires. Our data assimilation capabilities are becoming more advanced.

In short, we are further along in some of my ideas for how we can continue the evolution of CIMSS than I ever anticipated.

Is there an internal awareness issue that perhaps we should be addressing?

Maybe. Moreover, this may dovetail a little bit with how I view my role as director. I think people understand the research they are doing very well because they know their expertise and their team’s expertise. But I have a feeling that perhaps the interactions between groups within CIMSS might need to be bolstered. This extends also to the interactions with our external partners like AOS, NOAA, NASA, and DOE where interconnectivity and communication is key. They may have a grasp of what their team is doing, but I think there may be some gaps in that communication that we could look to address. It will be part of my role to make people more aware of the synergies between the various research programs within CIMSS and then, making sure that other units and agencies are aware that we have all of this expertise in-house.

How does your background in remote sensing mesh with CIMSS’ strengths? What do you bring to this mission?

One of the reasons I was excited about taking on this role in the first place is that I have been a user of CIMSS products and have known many of the CIMSS scientists from the remote sensing world for a long time. I think I am now in a position to actually start to put together the pieces – like we just talked about – and apply the talents of the various CIMSS researchers to some of these broader, more interdisciplinary projects. Returning to my own background, I obtained an undergraduate degree in physics and then pursued a Ph.D. in atmospheric science. I began my career working on satellite retrieval algorithms for cloud products, which is a huge part of what we do in CIMSS. Expanding the uses of the resulting datasets is where I feel that some of my expertise can really benefit CIMSS. For example, I think those datasets can play an important role in understanding aspects of the climate system that we do not understand or verifying climate model predictions. The same argument can be made for weather models – forecast models. We need to verify that the models are producing the right answer and we need to integrate as many observations as we can, into models, to improve forecasts.

These are topics that I have been working on for some time. My experience has been limited to a subset of remote sensing and over the last three months, I have been introduced to this entire team of outstanding scientists at CIMSS. I would say it is probably the highest concentration of remote sensing experts in the world, all of whom contribute to that broader goal of trying to improve climate prediction and weather forecasting for the public good. I think my expertise – and I am talking about the little research tasks that I have done throughout my career – is a good match for the types of research we could consider undertaking within CIMSS.

It is part of my vision to try to push even further along the path to be sure that CIMSS takes a leadership role in using the data products that we have been generating for so many years to improve climate prediction and weather forecasting.

Are there some other areas of emerging research that would be a good fit for CIMSS?

If we want to keep the thread of remote sensing going, remote sensing has evolved since the days of Verner Suomi and the launch of the first satellites. Our building has played a huge role in that evolution. The science has matured in terms of the way we exploit measurements to gain knowledge of the Earth but there are still many new directions in which we can go to improve the information that we gain from those measurements.

Here is one example of an area that some researchers at CIMSS have already been thinking about and it is something that is on my mind, too. When we retrieve a property of the atmosphere, we often look at a single satellite footprint and we apply some theoretical radiation calculations to it to try to determine what the satellite was looking at. However, if you think about the way humans interpret information, we actually gain much more information by looking at the broader context: looking at the spatial characteristics of a complete scene and how conditions have evolved over time. With the new geostationary satellites, GOES-16 and GOES-17, we now have 5-minute snapshots that can be stitched together to show very clearly how storms are growing. We now have a tremendous amount of information about the spatial context and temporal evolution at high resolution that, with very few exceptions, we do not typically use in our retrieval algorithms. We are still applying algorithms to each pixel of an image. I think the new area – and some CIMSS scientists are already trying to do this – will be to start to use advanced computational techniques, like machine learning, to try to extract information about entire systems as they evolve, switching our thinking to a whole system instead of its individual pixels. It is a much better reflection of the way humans interpret information.

We learn a lot more by looking at an evolving picture as opposed to looking at one single element. I see this as one direction, but there are several other research themes at CIMSS that could be considered emerging areas.

In reference to the different ways that humans interpret information, do you see new collaborations in areas of science that analyze how humans interpret information? Using that knowledge to inform the type of research we want to conduct here?

Indirectly, this is what the artificial intelligence (AI) community is doing. That is their exact goal: to take our understanding of how humans interpret information and convert it into a machine code or program so that we can actually train a computer to do the same sorts of things with data.

We are going to want to engage the AI community. Some scientists at CIMSS are already doing this, which is very exciting to me. I think in the next decade this is where remote sensing will head because of the vast amounts of information coming from new satellite technology. More and more we will engage and work with experts in the AI community to help develop those algorithms. Again, it is already happening, but will increase over the next decade during which time we may see it explode.

A slightly different angle on the question is that we have a unique capability right here in Madison due to the availability of a super computer and access to data. Few places have this.

If an institute is going to lead this effort, it makes sense that it is CIMSS.

Both SSEC and CIMSS have had leadership changes in the past six months. How do you see that organizational change in the context of the health of both institutes?

I sense a shared optimism about how refreshing this new enterprise could be. It is a big change for me, too, so that is part of my excitement.

Organizations need change to grow and thrive but when you take a new job, there is always the possibility that you have made a mistake or that you have gotten in over your head. You ask yourself, “Will I fail miserably?” I can say that over these last 100 days, that thought has vanished. Again, it is because of all of the conversations I have had with everyone in CIMSS – not just leadership, but everyone. The more I learn, the more excited I become about future prospects. I think that sums it up nicely.

Both Brad Pierce (SSEC Director) and I have talked at length about diversity issues within the Space Science and Engineering Center and CIMSS. We both are committed to trying to improve the diversity of ideas and groups so that everyone feels welcome here. This is not a unique organizational issue, and it is not specific to any group within our building. We know from research that the more diversity of ideas you have, the better the health of the organization and the better the range of possible solutions to problems. This benefits the community as a whole. I think AOS has made positive strides in the diversity area and I think this will benefit all of us.

I have been raising this topic in conversations. It is a something that we need to talk about openly, otherwise, we cannot change unintended bias where it exists.

In the end, it is all about the people. Because it is people who make organizations. The more I learn about the people and research of CIMSS, the more excited and optimistic I am for our future.

Do you have a mentor, or someone who has influenced your thinking or career trajectory?

I suppose I have had a series of mentors but I model my career after Graeme Stephens who was my Ph.D. advisor. Of course, he has given sound advice over the years, but I also just admire his level of creativity. His ideas are transformative. You can see over the course of his career how he has advanced science, advanced remote sensing and advanced the way we think about observations. In my opinion, many of the ideas represented in the new NAS decadal survey have his fingerprints on them. In terms of CIMSS and SSEC, we need people like that, like our founders, Verner Suomi and Bill Smith, who have been the architects of change in the field. Had I known Suomi myself, he would have been a mentor for me.

Actually, another person who worked directly with Suomi and who has influenced me, is Tom Vonder Haar. He graduated from the UW and went on to influence, and lead, the program at Colorado State University where I did my graduate studies. He, too, had that same ability to not only see what could be done today, but also think about the next generation of technology and its possibilities and trying to get the community ready for the next generation of sensors.

What have you enjoyed most about this transition period?

There is no doubt about it – it is all about meeting the people at CIMSS and hearing their ideas. I am stimulated by good scientific conversations. Every meeting I have had so far, and everyone I have met at CIMSS, has stimulated new ideas. I have been extremely impressed with what the scientists are doing and learning about all of their hidden talents. I had heard of, but had not met, many of them, so having these face-to-face conversations is critical. Hearing their ideas and seeing their reactions when I also share my ideas: that exchange has been so energizing.