A recent report from the Intergovernmental Panel on Climate Change (IPCC) confirmed that the effects of climate change are already being felt around the globe, leading to serious problems in agriculture, public health, and water supplies. “The one message that comes out very clearly is that the world has to adapt and the world has to mitigate,” said the IPCC chairman upon its release.
The design world’s increasing focus on resilience has made architects and engineers more aware of the need to prepare for an era of higher temperatures, rising sea levels, and other climatic changes. As the industry works to determine exactly what resilience means and how to achieve it, however, one of the biggest hurdles has been the lack of access to ready-to-use information about how conditions may change in a particular locale.
Arup recently partnered with climate consultancy Argos Analytics (whose vice president, Dr. William (Bill) Collins, is one of the IPCC’s lead authors) to create the WeatherShift™ application, which projects future weather patterns for specific sites based on climate models run for the recent IPCC report.
We spoke to Collins, Argos Analytics president Robert (Bob) Dickinson, and Arup’s Cole Roberts this spring.
*
Why is a tool like this needed?
Roberts: I happen to be sitting in San Francisco, and, like many cities in the world, it’s booming. There’s a tremendous amount of infrastructure and building design going on.
All of the design efforts in these cities rely on weather data that has been measured over time and averaged out for the purposes of predicting how buildings will perform and what kind of building systems are most suited to the climate. In addition to estimating energy demand and consumption, the historical data is used to estimate occupant comfort and determine life-cycle costs for business investment.
But all of the weather-based modeling is based in the past. Due to climate change, the industry is designing our cities and high-performance buildings, which will last literally hundreds of years, with outdated data. With the WeatherShift tool, we can finally start to take a legitimate, science-based stab at understanding future temperatures and weather conditions and use that to improve design.
Collins: We’re using the latest set of climate simulations from the most recent set of reports from the Intergovernmental Panel on Climate Change and the recent national assessment for the United States. So the foundations for the tool are the most current models, and the most comprehensive suite of those models, that are available.
Sample shift in natural ventilation suitability for San Francisco based on a pessimistic scenario (RCP8.5, 90%)
What’s the best possible scenario for how this could play out in the coming years?
Roberts: That everyone starts to design our cities, our buildings, and our infrastructure based on the climate they will reasonably experience, not the climate of the past.
Collins: I would completely agree. None of the students in my classroom have lived in a normal climate. People who are 20 years old live in a climate that is historically unprecedented since well before the gold rush in California. And these changes are going to get bigger with time. This is a chance to make cities and our communities more resilient to the changes that at this point we are certain are going to come.
How did the project come about?
Roberts: About a year and a half ago, Arup convened an international group to look at current best practices for how we model and predict performance of buildings using energy models — energy models which every single large building that goes up today, and even a lot of the smaller buildings, use. One of the outcomes was that we agreed we should find a way to start to design for the future climate.
This is a chance to make cities and our communities more resilient to the changes that at this point we are certain are going to come
Phil Haves at Lawrence Berkeley National Laboratory was at the meeting, and I said, “Somebody over there in your illustrious group of scientists must have a link to the global climate models. Can you introduce me?” He introduced me to Bill and Bob, and the rest is history.
Dickinson: We’d been doing some work on a pilot basis to help public agencies understand how various types of climate impacts would affect how they operate and how effectively they can execute their mission. That’s been our principal focus to date. But personally I’m very happy to see something getting out to the commercial sector as well, because I think that if society is going to deal with climate change effectively, we’ve got to have the private sector fully on board.
Collins: My interest in working with Arup really stems from the crying need to take what we’re learning about climate change and make it useful to people who have to make real-world decisions. Trying to understand how best to translate our findings into results that could help inform how to design better buildings has really been a great learning exercise.
Roberts: To some extent, we all live and work in our bubbles of expertise, and yet it’s the connections between our expertise where the real leaps of creativity and progress happen.
For example, when we’re designing high-performance buildings, one of the things we depend on is the ability of the buildings to cool themselves at night by using cooler outdoor air temperatures and then ride through the heat of the following day. One thing that we’ve learned from Bill and Argos is that the nighttime temperatures are actually going to be rising faster in many areas than the daytime temperatures. That’s a very important thing for people involved in the design of high-performance buildings to realize, because it means that reliance on nighttime cooling may not be nearly as viable in the future.
Sample dashboard output showing meaningful change in design summer and winter temperatures under the pessimistic condition (RCP8.5, 90%)
Collins: Nighttime warming is an important trend, and it’s actually one of the reasons why we have confidence in the projections — we’ve already seen this trend in the historical data. Looking back over records of temperatures in cities and around the country, one of the clear trends is that nighttime temperatures are warming faster than daytime temperatures. The diurnal range is becoming smaller.
Climate models are increasingly able to mimic the trends that we’re already seeing in real-world data. And the bottom line here, the challenge that climate change is posing, is that it’s taking existing patterns and amplifying them, sort of taking the volume knob and cranking it to 11.
Roberts: And it’s not just buildings that we’re taking about, but infrastructure at its most core for our society. For example, if we talk about warmer evenings, that means a potential increase not only in the amplitude of the demand on the grid, but also a shift into a different time of the day for those peaks to occur, because you’ve got both lighting coming on and a warmer evening. Plus you’ve got electric vehicles starting to plug in as people start to arrive home.
Climate models are increasingly able to mimic the trends that we’re already seeing in real-world data
The same thing goes for transportation infrastructure. You talk about the materials that are used in rail systems or highways, or even sidewalks and street curbs; in a lot of places in the country they’re designed for a certain temperature and climate. If those places start to get much hotter, you start to see failures in the horizontal infrastructure, not just the vertical buildings.
How will this application be used?
Roberts: Like all things, you never know exactly how it’s going to be used in the end. We’re open. We’re currently using it for building energy modeling: looking at the increased peak energy intensity, the amount of energy that’s consumed over time, and implications on system selection and passive design. We’re also using it for thermal comfort studies, both interior to buildings and exterior.
It’s likely we could be using it for microclimate studies, notably wind analysis, as well as precipitation and flooding events for site infrastructure work. And there’s no reason to think we can’t use it for portfolio modeling and city planning. Basically, anything that uses weather as an input to analysis for technical or strategic thinking is a possible tee-off point. Key is that the interest is in a solution that extends over the life of the city, infrastructure, or building.
Right now this is a proprietary data set, and we’re in a world of increasingly open data sets, or at least a desire to see open data sets. So it will be interesting to see between Argos and Arup — who are both very socially minded and impact-driven businesses — how we can find a way to make this information usable by a broader spectrum of the building industry.
How did you develop it?
Collins: The general idea was to figure out how to take the existing historical conditions that builders have used to figure out what a typical year looks like and shift that typical year into the future. We had to know not only how far to shift it — what years were we concerned about, how far into the future — but also what were we shifting: just temperature, or temperature and rainfall, or temperature, rainfall, and humidity?
So Argos and Arup considered what the relevant timeframes were, and also what information was really pertinent to building designers. We talked about how to take a fairly large ensemble of model stimulations and condense it down into a range of useable conditions.
The WeatherShift application can dynamically indicate changes in two common indices of comfort, wind chill (top) and heat index (bottom)
The other challenge that we faced is that we’re dealing with uncertainty. We can estimate how likely it is to get warmer by 2 degrees, 5 degrees, 1 degree. We had to fold that sliding scale of outcomes into the information that Arup and Argos jointly produced.
So the three pieces of information that went into how we designed this were the when, timeframe; the what, the information required; and then the how, which is how to deal with the sliding scale. Once we had agreed on those three factors, I went off and — to now get to the nerdy bit — downloaded all the information from the global archives of model simulations that have been assembled for the IPCC and global national assessment that are unrestricted for commercial use and built a code that compares the future against the present. And that provides the shift. I boiled down literally terabytes worth of information into a small set of these shift factors that I exchanged back to Arup. And then the Arup side of the story begins.
Roberts: Arup had some history of creating shifted weather files for a number of the cities in the UK. We took inspiration for the WeatherShift application from this prior work. We had had some of our analytical coding already in MatLab; we translated that over to a web-based application coding, improved it, and set up batch processing so we could call any number of the shift factor distributions contained in the Argos ClimateDelta™ dataset.
We’ve got tools to create the same shifts not only for the first 50 cities which we’ve done and published, but for literally thousands of cities around the world. We just don’t yet have the bandwidth to produce that, but if there was a call for it in industry I’m sure that both Argos and Arup would find a way to work with everybody on it.
Map showing areas of the world where the first 50 cities for which weather-shifted data has been released are located
What are the challenges that you foresee as the tool is rolled out?
Roberts: From Arup’s standpoint, I think one of the greatest challenges is finding a path that takes us from something that one firm is doing to something that everyone is doing. As large as Arup is at 11,000 staff, we likely design well less than 0.1% of all built area globally. If we’re convinced, as we say we are, that we need to act collectively and quickly as an industry — and even just as humans — we have to get the data out to the broader world. We’re looking for the right institutional partners that can help make that a reality.
There are very, very few engineering firms and architectural firms that are asking their clients how they want their buildings to consider future climate
And one last thought — the technical is often the easier piece; the legal and behavioral side is harder. Right now I believe that there are very, very few engineering firms and architectural firms that are asking their clients how they want their buildings to consider future climate. There’s a real question that needs to be answered: is the lack of conversation leaving the architecture and engineering profession open to legal claims? Based on rigorous model-based projections, the industry knows that there are changes coming. The extent of those changes is uncertain, but the industry as a whole could certainly be communicating better to their clients that it is something that is reasonably easy to anticipate.
In defense of industry, there is an equal shortage of clients asking the same question. Perhaps we all need to get better at preparing for the future.
Interview edited and condensed.