May 16, 2013 3:00-4:30 PM EDT
The presenters included: Rebecca Lupes and Rob Kafalenos of FHWA Office of Natural Environment, Mike Flood of Parsons Brinckerhoff, and Joel Smith of Stratus Consulting, Inc.
Welcome to the transportation climate change and extreme weather vulnerability assessment conference call. I will now turn the conference over to your host, Becky Lupes.
Becky Lupes, FHWA
Welcome everyone to the first webinar in the four-part series on transportation climate change and extreme weather vulnerability assessments. All webinars are being recorded and will be available on the website after the fact. Future sessions in the series will cover system-level vulnerability assessments and how to apply the results in decision making, and then we will have a special session on lessons learned from Hurricane Sandy.
In today's session, I will give a brief introduction to Federal Highway's vulnerability assessment framework, then we have a great lineup of experts who will delve into determining climate information. Mike Flood of Parsons Brinckerhoff will present on assessing criticality of transportation assets. Then Joel Smith of Stratus Consulting will talk about developing projected climate information for transportation asset analysis. Rob Kafalenos on my team at Federal Highway will give a case study discussion of our Gulf Coast 2 study and talk about sea level rise and storm surge projections. Then we will have a Q&A session.
Understanding how climate change effects and extreme weather will affect your transportation network is the key first step for climate change planning. To help with this, we at the Federal Highway Administration (FHWA) have developed a climate change and extreme weather vulnerability assessment framework. This webinar series follows the general structure of the framework.
The framework is two things. First, it is a suggested organizing mechanism for transportation agencies planning to conduct the climate change or extreme weather event vulnerability assessment. Second, it is a structure for us at Federal Highway that helps us to organize resources and lessons learned for reference and use by transportation agencies and others interested in any aspects of the work.
Here on the screen, you can see a graphical representation of the framework. Each box or circle has corresponding information, examples, and other resources that go with it and it's all contained in a 50 page document on our website that will be available for download at the end of the webinar. We will have a section of the website devoted to the framework where you will be able to click on the applicable part of the graphic and link directly to updated information and even more resources related to that bit that is included in the framework. We are thinking of it as a living resource.
The framework draws from the work and experiences of five vulnerability and risk assessment pilot projects that we sponsored in 2010 and 2011. The projects were conducted by the Washington State DOT, the Metropolitan Transportation Planning Commission, in partnership with the San Francisco Bay Conservation and Development Commission, the Oahu MPO, the Virginia DOT, and the North Jersey Transportation Planning Authority with the New Jersey DOT and the other NJ MPOs. The framework also draws examples from U.S. DOT's ongoing Gulf Coast Phase 2 study and work by other agencies around the country. As you can see, the framework is comprised of three main components: defining scope, assessing vulnerability, and integrating results into decision making.
The first component is defining scope. One thing we learned through the pilots was that it is helpful to define your project scope as much as possible at the beginning of the process. A central aspect of this scope is defining and articulating your objectives. What are you looking to achieve? One example from our pilots and in the framework is the San Francisco pilot team's vulnerability assessment work. It was part of the Adapting to Rising Tides Project, which was designed to bring the San Francisco Bay communities together to prepare for the impacts of sea level rise. The purpose of that project was to help the region's transportation planners improve vulnerability and risk assessment practices and to help formulate effective adaptation strategies. A goal of the project was to develop an approach for more fine-grained assessments. So, the team limited data collection efforts to near coast infrastructure, limited the study area to one county, and limited the project scope to analyzing the impacts of sea level rise.
Also, consider based on your objectives, what assets are relevant for your study. For example, the Washington State DOT pilot team focused on the transportation infrastructure it owned which includes roads, rails, ferry terminals, and airports. They included assets already in place or soon to be constructed as opposed to proposed projects in transportation plans. The New Jersey pilot team conducted a detailed analysis using a statewide travel demand model to identify zones with the greatest travel activity and used that as a basis for determining critical corridors and assets in the study areas. A little more detail on what they did and how various agencies selected their assets for study is available in the framework.
As much as you can upfront, consider what key climate variables you want to study. Consider past weather-related disruption. What climate variables are problems in your area? Consider the specific thresholds in which the problems start to occur. For instance, is there a specific temperature or duration of days above a certain temperature where the system begins to experience impacts?
A couple of examples from the earlier pilots: the Washington State DOT pilot considered sea level rise, precipitation change, temperature change, and fire risk. The San Francisco pilot limited its climate variables to sea level rise only and looked at storm surge, as well, with a focus on a portion of the East Bay potentially subject to impacts from inundation. They made use of sea level rise scenarios being considered by the state of California for adaptation guidance.
The next component, and the bulk of the work, is with the vulnerability assessment itself. Once you've defined the scope you can begin to assess the vulnerability of your assets. This isn't a linear process - information collected on assets may inform the climate data needs and vice versa. The framework includes resources for and examples of how you might do this.
In the webinar today, we will hear from Mike Flood, Joel Smith, and Rob Kafalenos about assessing asset criticality in developing climate input. In webinar two, we will hear from Washington State DOT, the San Francisco Bay area Metropolitan Transportation Commission, and North Jersey Transportation Planning Authority on how they used information on transportation assets and climate projections to identify vulnerabilities.
The third component of the vulnerability assessment framework is integrating results into decision making through identifying, analyzing, and prioritizing adaptation options and for incorporating assessment results into programs and processes. In the third webinar of the series we will hear from the Boston MPO and the Los Angeles County Metropolitan Transportation Authority on ways they are incorporating results of vulnerability assessments into their transportation planning processes. Some examples from our pilots that we discussed in the framework are the Washington State DOT, which incorporated the results of their vulnerability assessment study into project level and environmental guidance for the state. The Oahu MPO vulnerability assessment results ended up being used in developing legislation on incorporating adaptation into statewide planning.
The five pilot projects we had were focused on vulnerability assessments. I want to mention we are just getting underway with another 19 pilot projects and this time we are funding many that are focused on adaptation both from a programmatic standpoint and specific engineering options. Once those pilots are complete, we plan to make another big update of the framework perhaps terming it an adaptation framework rather than vulnerability assessment. That's what I had to talk about today, and now I will turn it over to Mike Flood of Parsons Brinckerhoff.
Mike Flood, Parsons Brinckerhoff
Thank you, Becky. Good afternoon, everybody. I want to first thank the Federal Highway Administration for inviting me to participate in their first webinar on climate change adaptation. It's a nice honor to have. I'm going to run through a short presentation today and the basic outline for the presentation is to define what we need by the term 'criticality,' talk about the reasons for thinking about critical systems, and discuss various methodologies for determining critical facilities as part of climate change adaptation practices.
The term 'critical' has two basic meanings for climate change adaptation, the first being that which is at risk, which is a piece that comes farther on in the process here. The components we want to focus on here are those components that are critical to the viability of the local region or state. The way we talk about it is that critical assets are not the ones that help you get out of the way of the storm but the ones that help your community function should an extreme weather event come through. What does your community need to do, have in place to function moving forward? Why do we need to think about criticality?
One of the reasons for identifying critical assets, and the one we used for the work on the Gulf Coast project, is to figure out a way to focus limited resources on studies or projects. The issue of criticality could also be applied to help set agency policies on risk tolerance. The basic question behind this is: are current risk policies, for instance a 100 year storm, acceptable for all components of infrastructure or should we be thinking differently about certain components of infrastructure that might have a higher importance and therefore, should they be damaged, have a higher level of impact on the transportation system? Like the two examples shown here, the bridge across Lake Pontchartrain, Louisiana as well as the South Ferry Station in New York.
There are a couple different methods to put in place to determine criticality. First, is a desktop method or some way of sitting at your desk and applying data to come up with a ranking or score for criticality across the network. Second, is a stakeholder method where you can sit down in a room with a group of folks that are well familiar with the transportation network to identify assets deemed critical. And the third is a hybrid approach where you apply assessment methodology and involve stakeholders to resolve the final determinations.
There are some drawbacks associated with each method. For the desktop method, you sometimes fail to reflect local input. And you can get some feedback from stakeholders or the community after you complete the desktop method that your method didn't quite capture their input, the community input. There is a lack of data for private transportation entities particularly on the rail and port sides, which makes it challenging to adopt a desktop type method. On the stakeholder side there is no basic assessment behind the work. This could be affected by personal preference like many transportation type projects. Also the facilitation methods can impact results. Those are some of the challenges to implementing the two basic frameworks.
I will run through a few of the criticality methods that were applied on some of the pilot projects to give you a better feel for what's involved. In the New Jersey project, the team identified criticality in terms of low to extreme based on a combination of population and job density, i.e. locations where most of the activity would be expected to occur. In Hampton Roads, a number of other varying criteria were identified -- assets that carry hurricane evacuation routes, assets of very high use, and assets that are already identified as priority runs or those at low-lying elevations. So this project used some combination of the two terms for criticality - what's most important and also what's potentially at risk.
For the Washington DOT, they did an assessment for their entire network of roads. They assigned a criticality ranking for each of the roads including basically from one to 10 and you can see across the bottom the factors that were included in the assessments and one of the ones to point out is whether there is an alternative route available. That's a factor in many criticality assumptions.
What I will walk through now is some of the work done on the Gulf Coast project. The Gulf Coast project was one that was sponsored by the U.S. DOT and managed by the Federal Highway Administration. It is being done in Mobile, Alabama with Kevin Harrison at FARPC. It involves what we term the Robs at Federal Highway Administration and also the good folks at ICF, the contract manager, and who we were doing the work for.
The project is in Mobile, Alabama so those of you not familiar with Mobile, it sits on the Gulf of Mexico down in the southern part of the country. What was the overall purpose of the effort was to try and determine the subset of the entire transportation network on which to perform the vulnerability assessments and identify the adaptive measures. Again, we were trying to determine a subset of the overall networks across all modes of travel to advance to later project phases.
What is critical infrastructure? How is it assessed? What we are trying to do is develop a consistent methodology across all modes. In terms of basic theory, what does critical transportation mean? It means that it provides connection to the broader network. How does the transportation network connect Mobile to the region and beyond? It has its purpose. It makes connections amongst multimodal centers and distribution centers. And it has functions - it has a societal function and provides access to employment, hospitals, etc.
In coordination with the project team, we came up with an assessment methodology that looked across three basic categories, which were socioeconomic, operational, and health and safety. On the socioeconomic assessment, we looked at the service to the regional economic centers, the availability of a redundant system, and whether or not it makes community connections. Basically, infrastructure that is important to the region. We looked at operational considerations. This is how often or how much the facility is used, its functional classification, whether it is a freight route or other factors that could affect its importance to the region. There was a health and safety element, this is a coastal area and there are early discussions on criticality identifying items like hurricane evacuation and disaster relief and disaster recovery are on the minds of many folks in the area. Also, the obvious one of whether it provides access to hospital facilities is an important consideration.
Before I talk about this slide, when I started showing it to people put their heads to the site to read. I want you to know that I will rotate the scoring up on the next slide so you don't have to look that way. Basically, in order to define critical infrastructure we wanted to delineate important assets and develop a scoring method to rate items for criticality and then the last piece being that we need to apply engineering judgment to fill in the gaps. What I've done is rotated the table so you can see it. One of the things to note is this is the result of both the desktop or stakeholder method. We developed a desktop methodology and scored and ranked various components of the modal systems. And then went back down and met with the stakeholders and they felt that the methodology did not quite capture all of their concerns in the operation of their social fabric in the area. One of the most important things being we show access to the shore areas into the fishing areas, which were showing up on the operational classification. These were identified as very important to the societal fabric down there. We added a row and then had them rank each one in terms of their importance. You can see we scored from 1 to 3 on how well each link in the network functioned in terms of other criteria and came up with the score at the end.
When we talk about system redundancy or whether there is redundancy available, there was a test applied. The region had a robust forecasting model that was applied. On the highway side, we were able to basically disable a link to represent potential impact of an extreme storm event and see what the impact might be on the surrounding network and try to determine whether the remainder of the network could function given the disabling of that system and what we tried to do is, for every link, we tried to select links that were representative of the network of the classifications of roadways and had the geographic coverage so we could understand the impact on the broader network.
I brought up another mode for the idea of showing you the challenge of identifying information for private stakeholders. The Mobile area is a huge multimodal area that has significant rail infrastructure going into and out of Mobile. It has port facilities and highway facilities and other facilities in there so we felt it was important to identify the critical rail infrastructure. We reached out to some of the operators in the area to find out information. For those of you who have done that before you know it's not that easy. What we did is we had one of our internal experts who knew the system pretty well come in and assess the infrastructure in the area and determine its criticality. In particular, in the private modes, the engineering judgment becomes very important.
Therefore, the results of the system and this was again the result literally of people looking over the fence at yards to understand their operation and the same thing, we scrambled field inventory teams for the port and the airport infrastructure in the area.
In terms of what Becky presented earlier on the Federal Highway assessment framework, you will see where this fits in. This is in the second phase, which is assessing the criticality of the system. You can see that it is listed as optional, and why would that process be an optional process? That's basically a development from the first phase to the second phase of Federal Highways' process. Why would it be optional? You may want to focus on all assets. For example, the MTC and San Francisco identified that they did not want to focus only on what they figured were high-capacity transportation options; they also wanted to focus their vulnerability framework on bike paths and other facilities that might not fall within the bounds of a normal criticality assessment. It can be a politically unpalatable process because you are taking a shot at telling other political officials or other people in your area what you identify as critical infrastructure and that can be a fairly challenging exercise.
There is more information available. These are two links to Federal Highway webpages which provide quite a bit of information and background on the criticality assessment process. The one link I did not provide was for the Gulf Coast 2 study which is also available through the Federal Highway website, which provides a direct link to the report that was issued in Task 1. There is some contact information, but you can get in touch with Becky or Rob or anyone on the Federal Highway team to provide more information on this. If you want to reach me, you can reach me through them. Thank you for your time today. I look forward to your questions later.
FHWA Vulnerability Assessment Framework
Assessing Criticality in Transportation Adaptation Planning - FHWA
Becky Lupes, FHWA
Thank you, Mike. Next, we will hear from Joel Smith of Stratus Consulting. He will talk about climate information for transportation asset analysis.
Joel Smith, Stratus Consulting, Inc.,
Thank you Becky and thanks for being here. It's an honor to join you and a pleasure to be with you. I will talk about some climate change information, in particular the use of scenarios, how we do it, and will give examples on different approaches and a lot of it builds on what you have just heard about.
The challenge of using information on climate change, there are a number. Basically, we cannot give a forecast of the exact changes in climate. We know the climate is changing, it's a fact that it is changing and we are highly confident there will be continued change, continued warming. The problem is we vaguely know how it will change. We know temperatures will go up, we know sea level will go up, we know there will be an increase in intense precipitation across the world. We know the direction of these changes. But we don't know exactly how these things will change because of uncertainties about emissions of greenhouse gases, the general response of the Earth's atmosphere to these changes, a number of factors. So we cannot make a precise or accurate forecast. If somebody says it's going to be 77 degrees today, well we can't tell you that in terms of the climate. That does pose a challenge. The question is, what do we do? Assuming the climate is going to be like the past, we know that's wrong so what do we do about going forward?
There are two approaches. One is called a scenario approach where you use outputs from climate models to develop a scenario: a plausible combination of conditions, something that could happen. Another approach is a threshold approach, what we call the bottom-up perspective, which uses output from the models and I think a lot of what Mike was talking about is really more of a threshold approach. I will talk about both beginning with the scenario approach.
We are using scenarios of climate change to put an envelope around the uncertainty. The scenarios should reflect ranges of different greenhouse gas emissions. Different scenarios of how much greenhouse gas emissions can change and to give you an example of the uncertainty of that, the U.S. was facing rapidly increasing emissions until about five years ago when natural gas started coming on and displacing coal and now our emissions have fallen. Whether they will keep falling into the future it's hard to say, but there's a lot of uncertainty about these emissions although in general they are expected to go up across the world. There are different changes in the Earth's climate. How much does the Earth warm? The standard thing scientists talk about is if we double carbon dioxide levels in the atmosphere. The best case is a 3-degree warming, that's in Celsius, so that's almost 5 degrees Fahrenheit, and then there is a range anywhere from 1.5 to 6 degrees, and that's the 90 percentile confidence level. That's a factor of 4 difference. There's a lot of uncertainty in how much the Earth warms for specific concentration of CO2 and then there are different changes in regional climate. Everyplace will probably get warmer but will it get wetter, not necessarily. Some places get drier. Some get wetter. We are confident the Northeast will get wetter, the Southwest drier; we're not as sure about the Southeast, the Pacific Northwest or the middle of the country. So there are those kinds of differences. Basically, we are trying to capture all of those or some reasonable range of those in the scenarios. The important point to communicate is the uncertainty and to understand the sensitivity of your system to different outcomes.
Here is an example of work we are involved in and Mike mentioned New Jersey Transportation Planning Authority's work; I'm going to show some examples from that. Here we picked three scenarios to capture three important elements of the uncertainty. If you look at the table, the first row shows low, we have an emission scenario and B1 may not mean anything to you. It is one of the emissions scenarios of greenhouse gases that the IPCC (Intergovernmental Panel on Climate Change) came up with around the year 2000 - that's a very low scenario, a very green scenario. We combine that with a very low sensitivity. What if the climate does not change much for this doubling of carbon dioxide? 1.5 degrees, and then we picked the model (MIROC from Japan) that did not show a lot of warning. So those were three things combined together that produced a low amount of warming. Then we developed the middle scenario (Mid) used an emission scenario called A1B, this one is approximately in the middle, it doesn't make it the best guess it just happens to be in the middle. Its sensitivity we assume the sensitivity of 3 degrees, and that is the most likely sensitivity for how much the Earth would warm for carbon dioxide and then we used what we call an ensemble, an average of all of the general circulation models. At the high end, we had a scenario called A2. You may have heard in the news the carbon dioxide level just reached 400 parts per million, there had been 280 parts per million. This high scenario would get us to between 800 - 900 parts per million by 2100. We combined that assuming that the climate is very sensitive and then we get 4.5 degrees of warming and pick the model, this is a Goddard Institute model developed by Dr. James Hansen, this happened to show a lot of warming for the New York region. We bracketed it in these three ways. We get a very low amount of warming, a middle, and a high amount of warming.
Here are some examples. Some of the things you can't look at are changes in average conditions, but you can look at changes in extremes. In this case, the bottom-up analysis, in NJ they are worried about the number of days above 95 degrees. They said that would be a problem for road surfaces and things like melting or softening of the pavement. This map shows under current climate roughly around 1990, how many days in the summer period would exceed 95 degrees. You can see it varies from about 2 to about 7. This is the high scenario where you combine the high emissions, high sensitivity, and a very high model and here you see it varies from about 55 days up north, 60 days in Moorestown, and 45 days down in the Southwest so we are getting far more days above 95 degrees at the end of the century than we had around the turn of the 20th to 21st century.
This map shows change in precipitation. This shows the change in the 100-year precipitation event so it happens once in 100 years, basically. It's a 24-hour precipitation event so here the concern is flooding. This map shows under the high scenario how much more frequent that precipitation event would be. We are getting it roughly -- the 100 year event could be as frequent as 25-35 years in the southwest part of the state, to every 14 years in the north part of the state (Long Branch) where you get very high frequency. So rather than once in 100 years it becomes once in a 14-year event. That gives you a sense of how much more frequent extreme events can become.
I want to give a little bit of a diversion here about some of these models and I would say if people are interested in finding out more about these models or where to get information, e-mail me and I will point you in the right direction. There are two types of approaches. The main information we get and what I've been showing comes from GCM, general circulation models that model the entire atmosphere. They divide it into grids and a typical GCM will have a grid that is roughly about 150 miles across. It's about 150 miles by 150 miles and will estimate a uniform climate for that grid. That includes a lot of variety. Oceans, land, inland areas, lakes, mountains. There are also approaches for getting 'downscaling,' more precise information. One approach is regional climate models, RCMs. These are models that have higher resolution. Their resolution might be only a grid box of 30 miles across. You have many more grid boxes and can better capture coastal areas, mountains, large lakes, those kinds of features. They also give a lot more heterogeneity. So if you look on the upper left, it shows a typical GCM projection of temperature and you see fairly smooth fields in terms of how much warming you get, the different colors represent different amounts of warming. Now if you look below that, this is from the same GCM, what it looks like in an RCM and you see a lot more regional resolution, a lot more specificity. The same thing goes for precipitation. If you go to the upper right, we see it drier along the West Coast; as we move inland, it gets wetter, but it's fairly smooth, and there are broad boundaries [between the specific temperature levels]. When you look at the RCM there is the general drier to wetter transition as you move from west to east but you see a lot more spatial heterogeneity and a lot more complications.
Is the RCM more accurate? Here's the problem. The different climate models, the GCMs, they are the basis for information we have on climate change. They don't agree on these patterns. So one model might show it's drier on the West Coast and then gets wetter as you go further east -another model may show a very different pattern. It might show wetter across the whole area. There are now 40 of these models out there that are being used for the intergovernmental panel on climate change. Each model might give very different patterns that on average there is some agreement. As I said earlier, the models tend to show drying in the Southwest U.S. and wetter in the Northeast, but not every model shows the same thing. The problem is these regional climate models have to -- the inputs are from the 'boundary conditions' and they get what's happening around the globe, and then they run the regional models, and so if you have a wet model, the RCM will divide up that precipitation precisely. If it is a model that predicts drying, it will divide up the drier conditions. You won't get the same output. You get a lot more precision, but it does not tell you what's going to happen. So be aware of that when you see this. The RCMs do not resolve differences between the GCMs, but they can increase the differences. This is a report (refers to a report entitled Options for Improving Climate Modeling to Assist Water Utility Planning for Climate Change from WUCA: http://www.wucaonline.org/assets/pdf/pubs_whitepaper_120909.pdf)
that some of us worked on looking at issues like downscaling. We found that downscaling will not reduce the uncertainties across the GCMs. RCMs may not correct the GCM errors but they do give insight into what can happen at a regional scale because they have more features; so in the long run they are promising but I want to urge caution in using them.
Another thing is looking at sea level rise and storm surge. For New Jersey, we looked at three factors; one was the change in global sea level rise and static sea level rise. There is uncertainty about that but basically it's thought it can be from two tenths of 1 meter to 2 meters by 2100. We used three scenarios: 50 centimeters, 100 centimeters, and 150 centimeters by 2100 to capture some of the uncertainty about sea level rise. The second bullet says regional variation. How much the sea rises at a regional scale varies because there are such differences and currents, density of seawater. You may have seen news accounts that the sea level rise along the East Coast is higher than the global average; that is picked up here. Then we add in storm surge. What we did was use the highest observed water level (HOWL) and that is a historic figure. Note: this is from 2011 so this was pre-Sandy. These are storm surges pre-Sandy and this is what they were in the table, and Sandy was much higher; I think it was 14 feet or more. We will combine these three sources of information. The one thing we are not doing here is estimating how the storm surge would change with climate change, and with higher sea surface temperatures.
Here's an example of what we show. This looks at Atlantic City. I don't know if you can see this clearly. If you look hard, the dark blue is areas that are already underwater. The lighter blue is what would be inundated by a 1 meter sea level rise. The red is when we add in the historic observed water level. If you look carefully, you will see a lot of land particularly in the southwest part of this picture that would be inundated by sea level rise alone. Along Atlantic City, a lot of the ocean part the storm surge would get them. If you had one and a half meters, you would have more land inundated, half a meter means less land inundated. But again, that's the way you can show what areas may be vulnerable to sea level rise alone as well as a storm surge.
But let's look at the bottom-up methods which Mike was talking about at the beginning. You start with an understanding of your system. In this case, the transportation system. What are you currently concerned about with regard to climate, extreme heat? Flooding? Storm surges? Intense rain events? Other things mentioned are critical thresholds and how much heat and for how long and how much rainfall would cause a problem? How much river flow or even snow, storm surge, all of those issues. And then we'd look at what is the current frequency for exceeding these thresholds? At this point we look at how conditions can change in the future. We've got those thresholds, and we come to the climate models and look at it and look at different scenarios. Each model output, each GCM or RCM could be considered a scenario. We might assemble a number of these scenarios. Maybe we have five or eight or 10 and we ask how many scenarios are the thresholds or tolerances exceeded? And when do they show this happening? So maybe we say we have eight models and six of the eight models by 2050 we get a certain amount of rainfall.
We want to be careful here because I noticed the likelihood of things occurring. The number of models showing exceedance is not the probability - it is just the number of models. For a lot of reasons, the models may be repeating errors; these are not purely random observations. You have to treat this information gingerly. But if you were to say one out of eight models shows we'd exceed a threshold by 2050, that's different if we say seven out of eight or all eight show to be exceeding a threshold. I'd be more worried about the latter situation. The advantage of the bottom-up is you focus on your system rather than the model output. You are understanding the sensitivities to climate and climate change, and you get some indication of how soon and with imprecise likelihood the thresholds could be exceeded. The advantage of the top-down approach is decision makers like to see it, in a lot of studies people give you the scenarios and do of all of our ability analysis. But you can also identify surprises, which you might not think about if you're doing a bottom-up approach. I think whichever approach is taken, the critical thing is to understand how your system could be vulnerable to changes in climate variability, and long-term changes in climate. I think it's important also this information gets used to begin working on contingency planning and measures to reduce risk. If I were to say, what should I use, bottom-up or top-down, try to use both. If you want further information, you are welcome to contact me (email@example.com) and thank you for your time.
Becky Lupes, FHWA
Thank you, Joel. Next up we have Rob Kafalenos of Federal Highways. He will talk about the Gulf Coast Phase 2 study.
Rob Kafalenos, FHWA
As Mike mentioned, I am one of two Robs you heard mention of earlier who manages the Gulf Coast project. Today I will talk about some of the results in the process we used to examine sea level rise and storm surge scenarios for Mobile, Alabama as part of the Gulf Coast 2 project.
The Gulf Coast Project is divided into two phases. The first phase provided an overview of climate change impacts on transportation infrastructure in the central Gulf Coast from Houston to Mobile. That study was finished five years ago in 2008. The phase two study is ongoing and is focused on Mobile. The first two tasks have been completed and they focused on developing the information to be used in the vulnerability assessment. Mike already talked about the work done on task 1 critical transportation assets, which he was involved in. The tasks in the second task include the work on sea level rise and storm surge which I will talk about today. The next two tasks focus on the vulnerability assessment itself and developing transferable tools to help our State DOTs and MPOs conduct these assessments in the future. This is a big study with a lot of different groups involved and some are listed at the bottom of the page. It's a DOT study and we have benefited a lot from cooperation with the South Alabama Regional Planning Commission.
Vulnerability is defined as a function of exposure, sensitivity, and adaptive capacity. This presentation focuses on one component and that is exposure: Exposure to sea level rise and storm surge. In the case of sea level rise and storm surge, the levels of exposure are represented by the inundation levels for each scenario we look at. We also considered two other climate drivers -- projected change in temperature and precipitation for three time periods over the century, but we won't cover that today. It's a whole other presentation in and of itself.
For sea level rise and storm surge, we selected plausible future scenarios. Scenarios are not predictions; they are tools to help us understand how transportation could be affected under a range of future conditions. We developed scenarios for storm surge and wave impacts using GIS tools to map the inundation of transportation assets under the various scenarios.
I will switch to talking about sea level rise and local sea level rise. In developing information on future sea level rise, we included both projected changes in global sea levels as well as historic rates of local land movement. For global sea levels, we selected three different scenarios: 30 centimeters by 2050, and then two for 2100 at 75 centimeters and one at 2 meters. Water levels rise as the volume of water increases in the oceans due to increase in temperatures and melting of ice. These scenarios were selected to cover the range of some of the more recent research on sea level rise.
Regarding subsidence, accounting for the vertical motion of the land is necessary for developing projections for local sea level rise. Rates can vary by location and vary greatly across the country. Some areas see subsidence of the land and others see the uplift of the land. For Mobile, we used a combination of benchmark surveys that were updated a few years ago. Also, satellite data were used to calculate historic rates of vertical motion of the land. The rates vary across Mobile, and benchmark survey showed a mean change of about 7.5 centimeters per century and the satellite data showed slightly less. Dolphin Island, which is this island you can see down here, showed slightly higher rates at close to 12 centimeters per century. In the end, we felt for Mobile, subsidence is not as big an issue as we expected. It is dwarfed by the change that could happen from sea level and from storm surge. In Gulf Coast 1, we found parts of central and western part of the Gulf Coast region, where the rates of subsidence could be quite a bit higher and in Galveston, Texas for example, the rates of subsidence were six times higher than in Mobile.
Next, I will show a few quick maps that show what areas were inundated through this process of looking at 30 centimeters of sea level rise. This map shows areas that could be affected by that level of sea level rise. You can't see much because there isn't that much affected. The next map shows 75 centimeters of sea level rise. This is the Mobile downtown area. This whole area is the study area. The study regime includes this area here, and some of the touchdown zones over here were some of the major pieces of infrastructure. The purple shows the critical rail facilities. The green shows the critical highway facilities. This is Mobile downtown and Mobile docks and port areas. The next graph shows the level rise with 200 centimeters with potential inundation, you can see quite a bit more potential flooding in these areas.
This is a summary of the results for looking at sea level rise. This table shows the percentage of critical assets, so these are the most important assets identified in the first task across each mode of transportation. It shows how much can be affected by each of the different sea level rise scenarios. We can see the 200-centimeter sea level has much greater impact than the lower levels. If we look at highways, we can see there is a huge difference. At the 30-centimeter level, we can see 1 percent of highways can be affected. That rises to 2 percent under the middle scenario, and the higher scenario of 2 meters is quite a bit more. Generally, you can see rates of increase across the different infrastructure categories.
Next, I will talk about storm surge scenarios that we developed for the study. For storm surge, we developed 11 scenarios and modeled them using the advanced circulation model or ADCIRC. The wave model was used to assess wave heights. We modified the historic storms to adjust the track of the storm, to raise sea levels in order to make it more intense to account for climate change projections. Why did we use historic storms instead of creating synthetic storms? Using historic storms increases the credibility of the work and provides a reference for communicating results. Especially for people who live in the area, talking about historic storms and using those storms can help communicate results to them and others who lived through those events. It also allowed us to use data available for each of those storms that had been recorded during those events. In this study, we used ADCIRC, we used it in part because it was an accepted model for developing inputs to feed into the design processes for highway projects in coastal areas. ADCIRC is more refined and closer to the shore and includes conditions further out and earlier in the development of a storm. Other processes or models such as SLOSH are considered appropriate for assessments for planning studies.
Next, we will talk about storm surge scenarios. We picked two storms to understand the range of potential coastal storms facing Mobile. The first is Hurricane Georges. It was a category 2 storm that hit the Gulf in 1998. The Gulf gets hit by category 1 or 2 storms every five years. Katrina, the second storm we looked at, was a much larger storm that hit the Gulf in 2005. It was category 3 at landfall and a large storm that caused considerable damage to Louisiana and Mississippi. Georges provides a reference impact for a moderate storm combined with sea level rise. Katrina and the permutations we developed allow us to look at the impacts of larger and less frequent storms. For Georges, you can see we adjusted the different levels of sea level rise. We looked at all three of the scenarios, 30, 75, and 200 cm. For Katrina, we looked at 75 centimeters and we adjusted the path of the storm and the intensity of the storm as well, to cover some potential effects. I will go through some maps that show the results just for the ones highlighted in black.
This is the first one, a model of the Georges, no changes in sea level rise were added to it. This is the lowest storm scenario we looked at. You can see some effects, not a lot. On the right you can see a gauge that shows the colors, the light color is at the top which shows less than 2 meters deep and it gets higher than that on the scale on the right. Next, we look at Hurricane Georges at 200 cm of sea level rise, you can see quite a few more areas are covered with the pink to red color. In this case, most of central downtown is flooded. Next we looked at Hurricane Katrina and this is Hurricane Katrina with no adjustments. Inundation is slightly less than Georges with 200 centimeters.
The next thing to talk about is adjusting the path. In doing the studies, the hurricane modelers tell us that no one path is likelier than another. We see now - this is the actual path of Hurricane Katrina. You can see it came in through the Gulf and hit just to the east of New Orleans. In the study, we adjusted it slightly so it would come in just to the west of Mobile. Because the winds rotate counterclockwise, shifting the path of the storm to this area to the west of Mobile Bay increases the storm's impacts because it pushes water up the Bay. This is the Katrina shifted path. We have not added sea level rise to it yet but you can see Katrina shifted shows greater levels of inundation than the actual Katrina storm. Surge depths at Mobile docks which are in this area, are at 19 and a half feet, which is 7 feet higher than an unshifted scenario. Wave height -- If you look at the changes in the color blue, the darker blue is higher wave heights and lighter means lower. The barrier island south of Mobile Bay here and here help reduce the waves that go in. A few more scenarios, this is Katrina shifted like we were looking at before. This is what 75 centimeters at sea level rise. It shows a gradual increase.
These last two scenarios are different than the previous ones. The goal is to see what could happen under more extreme scenarios. This is Hurricane Katrina shifted with 75 centimeters of sea level rise and on top of that including some effects of climate change on the storm. It shows reduced central pressure of the storm which is thought to lead to higher winds and higher wave heights. Under this scenario, the central pressure is reduced, with intensity increased. This caused the model storm surge to increase by about 2 feet to nearly 25 feet at the Mobile docks.
This is the last map and under this scenario, the maximum winds of the storm were capped as it went through as it hit landfall. The maximum winds of the storm were held at 150 knots and this caused the model storm surge to rise to 27 feet.
In closing, this is another graph or table and this focuses on the effect of storm surge on the different assets and it shows the different inundation levels. We cover roads, rail, ports, transit facilities, and airports. We also looked at pipelines, but we did not include them in this chart. For each of these, it shows the miles affected or in some cases the number affected, like the number of ports or transit facilities or airports. The numbers in blue are the percentages of critical assets that are affected. You can see for the critical roadways, it shows the greatest amount of variation by scenario from 16% to 58% from the lower end to the higher end. For some of the other types of transportation infrastructure, you can see the range of effects was not as great.
On the ports, they are closest to the water, so it is not surprising the ports show among the greatest levels of inundation. The proportion affected ranges from three quarters up to 96 percent for the ports. Similarly, freight rail, which is also close to the water, is also quite affected. More than half of the port facilities, 74 percent of the port facilities, are affected, and then for rail you can see more than half of the rail facilities are affected on the lower end storm surge scenarios and all the way up to 78 percent under the maximum.
We are working on the third task of the study and that is working on the vulnerability assessment that pulls all this information together. Thank you.
Becky Lupes, FHWA
Thank you. I would like to open it up for questions. You are welcome to ask questions in the chat pod or the operator will come on and tell you how to ask questions via the phone.
Q: Was this recorded?
Yes, this was recorded and it will be posted on the Federal Highways website. I will try to put a link to our website that we will have it posted on the web site shortly. Probably a few days after the webinar will be my best guess. Link: http://www.fhwa.dot.gov/environment/sustainability/resilience/webinars/
There are no questions in the queue at this time.
Q: Please explain subsidence as an issue in vulnerability assessments.
In some areas subsidence is not much of an issue. There are certain parts of the East and West Coast where there's not that much subsidence. Certainly in the Gulf Coast region it is something that needs to be looked at and there are areas around Alaska that it needs to be looked at and the reason we have to account for subsidence is it is a factor that goes along with whatever the assumptions are about future rises in sea level or future impacts from storm surge. Those things need to be considered together. If the land is subsiding, it can magnify the effect of higher sea levels for storm surges from existing storms. If the land is having uplift which is the case in some areas it to moderate some of the effects from storm surge in sea level rise. Great question.
Q: Could you talk about assessing criticality of pipelines?
I will take the next one about the critical pipelines. Critical pipelines was one we had to rely on engineering judgment. One of our engineers looked at it and basically the system was developed on the volume of individual pipeline facilities, and what was in them, what connections were made beyond the region, and whether they made connections to places like power plants or major commercial facilities in the region. That is the basic framework for how pipelines were assessed in the region.
Q: Have any of these studies resulted in actual transportation improvements?
The studies were just completed so probably not at this point, all at planning level. Though in the Washington State DOT study, they ended up with a map of vulnerable facilities. They incorporated it into their environmental guidance, their State NEPA guidance, so if they are going to work on a project, or have a study on a project that has been identified are on a facility identified as vulnerable, then they need to look at climate change concerns as part of their project development process. So they are incorporating results into project development. I'm not sure about any other specific projects at this point because these are all at planning level.
I should mention transportation infrastructure is being adjusted for climate change. In the Gulf, they rebuilt after Katrina, they built some higher. People are looking at improved defense following Sandy. There's a lot going on in terms of looking at improving infrastructure and making it more robust against climate change. I think a lot of this climate change is being taken seriously whether as a result of a study, these studies help contribute to the knowledge base to inform these decisions.
Q: Did any of the studies integrate the results into decision making?
There is a question about if studies integrated results into to the decision-making that the state DOT did. I know Washington DOT is using their results -- the third webinar we will have in the series is going to be talking a lot more about examples of incorporating results into decision-making. But at this point these studies were completed in 2011, two years ago, so I believe the results are going into the decision-making but the planning process is long. If they haven't updated plans, then maybe it's not included yet. But as far as I know, they are all being used.
One other point is we are talking today mostly about six or seven different studies, but there are quite a few other studies that have been done around the country, other areas considering vulnerability in doing assessments of vulnerability beyond the ones we are talking about today.
Q: When determining scope is project level better than corridor study?
I think it depends on what the objective is. Are you interested in looking at a specific project or are you looking at interested in getting a broader look at vulnerabilities?
If I can add to that, it relates to one of the questions I see coming up about the models. When we were working with the New Jersey Transportation Planning Authority and I forgot to mention Cambridge Systematics was the prime on that, New Jersey wanted to look across the system. The models themselves, you can downscale to the project level but to be honest it's more reliable in a broader regional scale and it's not a bad place to start. But you can still use that information to look at effects across a corridor or regions or what it might mean for an individual project.
One of the keys at the system versus project level is if you walk through the steps of the process of thinking about what are the bounds, what the potential might be and what are the risks associated with each of those levels and do I want to invest in addressing that risk or is that risk too much of an unknown for me or basically what is the increment that it would cost me at the project level to obtain a higher level of risk avoidance. I think it's a very similar process but it could be applied at either level, corridor or system level.
Q: Thoughts on migration of barrier islands with change in sea level.
There is a question talking about thoughts on migration of barrier islands with changes in sea level. I'm not an expert, I would say one would expect to see sea level to cause barrier islands to migrate and change more quickly than they otherwise would. I think one has to be concerned if the rates are too fast then it could overtake what the barrier island could do but these are questions that probably depend on how fast things change.
Q: What are the prospects for greater precision and accuracy in the models? Are the ranges of outputs likely to be narrowed over time?
I can respond to the next question on the prospects for greater precision and accuracy in the model. These models have been around since the 1960s. They've gotten better over time as computer power has increased and data improved. The models are getting better. The resolution of the models is improving a lot and at some point, the global models will be run at 1 degree which is 60 miles across and that's a lot of grid points when you think about the entire world. I think they are doing a better job of simulating current climate. We do see improvements over time. There have been studies that show models on average are better simulating current climate, so there is hope. One of the problems is whether we are getting more agreement? And that's the second part of the question. That has not been narrowed over time even as the models get better, we still get wide ranges of projections of whether a region will get wetter or drier or how much hotter. Unless there's some big scientific breakthroughs and understanding of world climate, it's unlikely we will see a narrowing of those projections, and in my presentation, there was a report that the Climate Alliance asked for to precisely look at these questions. The answer was we will have to live with the uncertainties for the time being, though we do have a lot more model information, and we can use that to better understand ranges of outcomes and get more insight into what might happen, but unfortunately the answer is no. But wait five years and look at the models agreeing, I would say no. If you need to make decisions now on adaptation, you will have to work with what you have basically is the bottom line.
Q: What effects does harbor dredging have in reducing storm surge impacts?
I don't have the answer to the Volpe Center harbor dredging question. Not sure what you're talking about in terms of dredging. Making the water deeper, I would think maybe Rob can talk about the impacts of deeper water on wave energy once it comes onshore in less deep water, but in terms of Brian Leslie's question about flood depth to use to assess damage caused to transportation infrastructure. On the Gulf Coast 2 project, this whole process comes together where we are trying to take downscaled climate data and methodologies to break it down to the engineering level and engineering assessments and benefit cost assessments to try and see how you would walk through one of these things on the actual project level. That part of the study should be completed by the end of this year I guess with some type of publication in 2014. That is the end phase of the Gulf Coast 2 project that Rob was talking about earlier.
I'm glad you raise that; one thing we did not do in the first Gulf Coast study, the one ended in 2008, is we did not deal with costs to any appreciable degree. The part of the study that Mike is talking about is an important part where we'd look at doing an in-depth analysis and looking at cost and doing cost benefit analysis.
To get back to the harbor dredging, Boston Harbor Association did do a recent study that you could Google and I imagine they probably looked at dredging but I'm not sure. Also, one of the new pilots we have going on is being conducted by Massachusetts DOT and they are looking at impacts on the Central Artery and potential solutions, so perhaps they will look at dredging.
One thing to say about changing depths, we do know in terms of coastal modeling when coastal engineers do the kind of modeling that they do, they have a factor they referred to as depth-limited waves, as the depths increase which could be from a variety of factors that can lead to higher wave heights. If you are looking at a particular dredging situation and everything, I imagine there could be all sorts of other local factors coming into play, in terms of how dredging would actually affect storms.
Question: Do you have knowledge of transportation projects where climate change adaptation has influenced the selection of alternatives for a given project, based on the vulnerability of the area? What about modeling/projections and its uses in the project development process and the NEPA process?
The next question is directed toward Becky.
I'm pretty sure in the New Jersey study; I think Dewberry calculated flood depth inland. We gave them change in precipitation. They looked at inland flooding as well, and of course as mentioned in my presentation, we used storm surge to look at essentially flooding levels, flood depth in the coastal areas.
I know that Caltrans has some strong guidance on sea level rise that needs to be incorporated into the transportation development process. I know from some conversations that North Carolina did some work on climate change for one of its coastal bridges, and Maryland has also started to assess sea level rise at least as an implication on bridges to barrier islands. So those are two projects where work is ongoing on thinking about this process.
In addition, I'm glad you mentioned California, also Army Corps has policy out in terms of sea level rise.
There are NEPA implications and that's one of the realities is that part of the decision-making process if you are going to construct a facility to address climate change, that might have implications in larger environmental impacts, particularly in coastal areas. So it could potentially be a challenging dialogue to address long-term climate change as part of transportation decision-making and how that fits into documenting the potential environmental impacts.
A couple of projects in Washington State as well have looked at potential climate change impacts as part of their NEPA studies and I don't know if it ended up changing the preferred alternative but definitely, several states do consider and look at potential vulnerabilities and impacts from climate change.
Question: Is the next Intergovernmental Panel on Climate Change (IPCC) assessment expected to provide new or better information for modeling and climate predictions?
Yes, they will be new. It depends on what is meant by better. As I mentioned earlier, the models do keep getting better, higher resolution, the physics is better, we have better observations. Just as an aside, these are the same classes of models used for weather forecasting and I remember hearing an interesting analysis of Sandy. The British forecast - that turn that Sandy made in the Atlantic and made a direct hit on New Jersey, which really was one of the major reasons it was as destructive - they forecasted that a week in advance. That would not have happened 10 years ago. There are better models, better data, these things are doing a better job of simulating weather and climate. But unfortunately, if you also ask, will they provide more agreement? From what I understand of these new model assessments; no, there is the same level of disagreement. It is a new set of models and I think they represent better science than what was in the older set of models. The IPCC, the Intergovernmental Panel on Climate Change is an international group of scientists put together by governments around the world and the World Meteorological Organization and they will be issuing their fifth assessment report this fall on the science of climate change.
There are a couple of questions on whether Federal Highways has formal guidance on incorporating sea level rise in project design. We do not have formal guidance. We agree that you probably should if you have that information.
That's a good point. The guidance we have issued thus far has to do with funding. We allow places to do different things with our federal funds. We put out a Federal-aid memo last fall, a joint memo with the Office of Federal Lands, Office of Infrastructure, and the Office of Environment &Planning, that talks about the different kinds of funds that can be used to conduct vulnerability assessments to consider climate information and project development, but as Becky said, at this point we don't have specific guidance on how to incorporate sea level rise in project design.
There is a question about a public EIS that addresses climate change and the impact analysis. You can look at recent ones from Washington State and California; both include sea level rise analysis and climate change impact analysis. New York State EISs as well, many of themdo; New York State's alll look at greenhouse gasses, and they sometimes look at climate change impacts. I suggest looking at them. There's the Columbia River Crossing in Washington State or some of their larger projects, State Route 522 in Washington State. Just yesterday or two days ago the Government Accountability Office put out a study on how Federal agencies are addressing climate change impacts, and they investigated Federal Highways as one of the agencies and specifically they looked at Washington State and Louisiana and one of the projects they highlighted was State Route 522 in Washington State.
There is currently guidance for designing transportation facilities in the coastal environment called HEC-25; we have an effort underway to put information together on how to deal with doing modeling for storm surge and sea level rise [and add this to HEC-25]. That is ongoing.
I can add a comment or advice on that. The National Research Council just put out a report because the West Coast states asked about sea level rise and they recommended using between half a meter and 1.5 meters. The Gulf Coast study used up to 2 meters. A lot depends on how risk averse you are. Do you want to avoid the worst-case; but that could cost more. Do you want to spend less and take risk? Can you do adaptive management and come back and do more later? I also suggest using multiple scenarios. Using one number can be misleading. You want to see, at least test how your responses hold up under a wide range of scenarios. That is critical in doing this kind of analysis. If somebody says this is the best estimate, don't just use that estimate - test for a reasonable range of potential changes.
Maybe I wasn't paying attention but I see there is an opportunity to download a bunch of technical reports. Is that available to everybody?
Yes. There is a file share pod on your screen and we made available the final reports from the five pilots that we helped sponsor, and also the framework document. Those are also available on the website. Click on them and save them to your computer.
Any other questions? No questions on the phone.
Thank you everyone for participating today. If you have any further questions there is contact information on the screen as well. Here is some information about the upcoming webinars. The next one Session 2 will be on system-level vulnerability assessments. And then we will have Session 3 applying the results. Session 4 will be a webinar on Hurricane Sandy and lessons learned. There is a link to register I believe. These are have been pretty popular and they might all be full, but you can try to register and if you cannot get in they will all be available on our website as reported. Thank you for joining us and have a great day.
Link to upcoming webinars and registration: http://www.fhwa.dot.gov/environment/sustainability/resilience/webinars/
That concludes the conference for today. Thank you for your participation.