DESIGNING PORT INFRASTRUCTURE FOR SEA LEVEL CHANGE: A SURVEY OF U.S. ENGINEERS

Seaports are particularly vulnerable to the impacts of climate change due to their coastal location. With the potential threat of up to 2.5m in sea level rise by 2100, resilient port infrastructure is vital for the continued operation of ports. There are strong economic and social incentives for seaports to provide long-term resilience against climate conditions. For example, service disruptions can cost billions of dollars and impact the livelihoods of those who depend on the port. Engineers play a pivotal role in improving the resilience of ports, as they are responsible for designing port infrastructure that will be adequately prepared for future sea level change (SLC). However, incorporating SLC is a challenging task due to the uncertainty of SLC projections, the long service lives of port infrastructure, and the differing guidelines and recommendations for managing SLC. Through an online survey of 85 U.S. port and marine infrastructure engineers, this research explores the engineering community’s attitude and approach to planning for SLC for large-scale maritime infrastructure projects. Survey findings highlight the extent that projects incorporate SLC, the wide range of factors that drive the inclusion of SLC, and the numerous barriers that prevent engineers from incorporating SLC into design. This research emphasizes that traditional engineering practices may no longer be appropriate for dealing with climate change design variables and their associated uncertainties. Furthermore, results call for collaboration among engineers, port authorities, and policy makers to develop design standards and practical design methods for designing resilient port infrastructure.


Introduction
Maritime infrastructure such as wharves, docks, and piers are particularly vulnerable to sea level change (SLC) due to their coastal location (Asoriotis and Benamara 2012). Observational data and calculated predictions confirm that sea level is changing (NRC 2012;Parris et al. 2012;IPCC 2013), and therefore, engineers of port and maritime infrastructure projects need to design structures to be more resilient by considering SLC throughout the design process (Esteban et al. 2011;Becker, Toilliez and Mitchell 2015). However, this can be a challenging task due to the uncertainty of SLC projections as well as differing guidelines and recommendations for managing SLC, especially at the local level (Becker, Toilliez and Mitchell 2015).
Furthermore, incorporating SLC considerations in port engineering structures is especially critical, as these projects tend to have long working lifespans, in some case exceeding 100 years (Becker, Toilliez and Mitchell 2015). There are strong economic and social incentives for seaports to provide long-term resilience against climate conditions (Becker, Toilliez and Mitchell 2015). For example, service disruptions can cost billions of dollars (Haveman and Shatz 2006) and impact the livelihoods of those who depend on the port (Becker et al. 2013). Despite the need for more resilient port infrastructure, there is currently no standard nationwide guidance for incorporating SLC information into design (Toilliez 2018). Therefore, if SLC is to be incorporated into a project, engineers must make subjective decisions on what SLC information they will use and what guidance they will follow.
To better understand how different firms, organizations, and individual engineers incorporate SLC information into a design, the researchers conducted an online survey of 85 port and marine infrastructure engineers. This exploratory survey addresses the following questions: 1. In what capacity are port infrastructure designers incorporating a sea level change projection into their design specifications for large-scale port engineering projects? 2. Where do incentives and disincentives originate for US engineering firms to incorporate sea level change into the design specifications of large-scale port engineering projects?
3. For engineering firms that are incorporating sea level change, what strategies are the port infrastructure designers in those firms implementing in the design specifications of large-scale port engineering projects to cope with the scientific uncertainty of sea level change?
By conducting a first-of-its-kind assessment of the current level at which engineers consider SLC in the design of port and marine infrastructure, points of intervention can be identified where collaboration can occur to effectively promote better resilience strategies. The baseline data resulting from this research can also be used for tracking how engineers change their approach to incorporating SLC into their designs over time.

SLC threatens maritime infrastructure
Globally-averaged, near-surface temperature records from 1850-2016 provide undeniable evidence of a long-term warming trend (WMO 2018) while temperature and salinity data from 1958-2014 suggest increasing ocean heat content (Dieng 2017).
As the temperature of the ocean increases, thermal expansion causes the average global sea level to rise (Dieng 2017). Recent scientific projections of global mean sea level rise (GMSLR) range from 0.3 to 2.5 meters by 2100 (IPCC 2013;Jevrejeva 2016;Sweet et al. 2017). However, regional and local scale SLC is less understood (Gregory et al. 2001). For example, SLC predictions vary widely across coastal regions of the United States. Parts of Alaska could witness a sea level decrease due to land uplift, while Louisiana may suffer a rise that exceeds global mean sea level rise projections.
The rate of SLC is uncertain because it is dependent on future greenhouse gas emissions (Church et al. 2013) and the complex mechanisms that control changes in sea level such as inputs from glaciers, changes in land water storage, and coastal erosion (Rahmstorf 2007;DeConto et al. 2016). The uncertainty of GMSLR is relatively minor over the next few decades (2040-2060), but increases substantially around 2080 (Church et al. 2013). These uncertainties need to be appropriately and transparently accounted for when planning for coastal hazards such as SLC (Stephens et al. 2017). Conversely, neglecting uncertainty about SLC projections can result in considerable underestimation of flood risks (Ruckert et al. 2017).
Infrastructure development and risk management decisions often come with long-term commitments that can be climate sensitive (Hallegatte 2009). For example, the engineered design life of port infrastructure is typically 30-50 years, but these structures will often have service lives that exceed 100 years (Becker, Toilliez and Mitchell 2015;Taneja 2010;UNCTAD 1985). Thus, many structures built today will have to cope with uncertain climate conditions in 2100. However, a survey of port administrators (Becker et al. 2012) found that capital planning cycles at ports are typically only 5 to 10 years. This mismatch between planning and infrastructure lifetime presents a concerning outlook on the ability for port infrastructure to adapt to a changing climate (Becker et al. 2012). Alternative design options are often investigated for port capital planning projects, but often the low-cost alternative is opted for, which can be attributed to difficulties in planning for uncertainty (Taneja 2010). Remarkably, a follow up port infrastructure investment survey in 2016 found that U.S. ports and their private sector partners plan to spend nearly $155 billion in port-related improvements (AAPA 2016). This shows that infrastructure spending from 2016-2020 is expected to triple that of 2012-2016. While portions of this investment will be dedicated to dredging and navigational improvements, the 2016 survey found that key investments are being planned for terminals, berths, piers, equipment, expansion, facility rehabilitation, and road and rail connections (AAPA 2016). Another recent study shows that $49 million at the Port of Houston has been invested to expand a container yard that will add 50 acres of storage; $36 million has been invested at Washington's Port of Everett to make the port ready to receive more cargo brought in by larger ships; and several others including the Alabama Port Authority and Port of Wilmington are engaged in dredging plans to bring in larger ships, which will lead to more cargo, and ultimately lead to the need for port expansion and infrastructure investment (Nabers 2018).
Seaports and port infrastructure will be especially vulnerable to SLC because they do not have the option to relocate, as their functionality depends on their coastal location (Asoriotis and Benamara 2012). Researchers predict that rising sea levels will affect 79 European ports by the end of the century (Christodoulou et al. 2018).
Officials from the Port of Virginia are expecting a foot and a half of sea level rise in the next 30 years, causing them to invest in raising electrical power stations and moving data servers farther away from the water's edge (Phillips 2019). Changes in sea level will also have a direct effect on other coastal hazards such as storm surge (Neumann et al. 2015). Therefore, seaports will need to make decisions about how to adapt to future climate conditions. Successful port adaptation and resilience strategies will require collaboration across various stakeholders from engineers, planners, financers, insurers, scientists, port operators, shippers, regulators, and emergency responders (Becker et al. 2013). Each group of stakeholders has a role to play in port resilience and adaptation, and each face their own set of challenges to overcome.

SLC uncertainty challenges engineers
Engineers and designers play a key role in increasing the climate resilience of seaports. They must consider not only their clients' needs, but also the needs of other stakeholders who depend on the infrastructure they design (Becker, Toilliez and Mitchell 2015). Since ports provide both private-sector profits and public services, stakeholders vary widely from shipping companies and insurers to local governments and local residents (Becker et al. 2013). The inadequate design of port infrastructure can have consequences for these stakeholders including physical damage to infrastructure and indirect damage to supply chains of goods and services (Becker et al. 2013). For engineers, the long service lives of port infrastructure coupled with uncertainty of regional SLC over the expected life of port infrastructure makes providing professional advice more challenging. Furthermore, a change in sea level can also have an effect on the geomorphology of a site, adding to that uncertainty (Becker, Toilliez and Mitchell 2015). SLC uncertainties, along with uncertainties of land cover and use, resource availability, and demographics in population in the future will require flexibility in infrastructure location and design (Olsen 2015). However, the standards, codes, and regulations that govern infrastructure are often slower to adapt to changes (Olsen 2015). This adds to the difficulty of answering the question: What level of SLC should engineers design for? Currently, there is no standard nationwide guidance for answering this question (Toilliez 2018) and there is a lack of design standards to guide this decision. There are several agencies and organizations such as the U.S. Army Corps of Engineers (USACE), National Oceanic and Atmospheric Administration (NOAA), and U.S. Department of Transportation Federal Highway Administration (FHWA), as well as local and state governments that provide their own guidance, but use differing scales, projections, and uncertainties of SLC.
In planning, there is always a gap between what is known and what should be known, and in order to bridge this gap, flexibility is needed (Faludi 1977). However, engineers have traditionally operated under the assumption of stationarity, which assumes that natural systems fluctuate within an unchanging envelope of variability (Milly et al. 2008). Stationarity can be used to analyze numerous environmental hazards and design variables. Using flooding as an example, historic flood analysis can provide an estimation of the extent and intensity of flood scenarios and associate an exceedance probability to it (Merz and Thieken 2004). The usual procedure is to apply a flood frequency analysis to a given record of discharge data (Stedinger et al. 1993) and to transform the associated discharge to defined return periods (e.g., the 100-year event) with an estimated inundation extent and depth (Apel et al. 2009). The error associated with these estimations is acknowledged, but assumed to be reduced by additional regional and observations data (Milley et al. 2008). However, these natural systems no longer exist in an unchanging system because anthropic impacts have caused the Earth's climate to change (Milley et al. 2008). Studies have shown that even small changes in climate may result in large changes in storm events (Knox 2000). These climatic changes led Milley et al. (2008) to declare that stationarity is deadit should no longer serve as a central, default assumption in risk assessment and planning.
This calls for a paradigm shift in planning, engineering, and design.
Furthermore, to increase resilience in the built environment, new adaptive planning and design strategies will be needed (Ahern 2011). Coming up with a successor is a daunting challenge due to complex patterns of change, significant uncertainties, and a constantly changing knowledge base from continual climate observation (Milley et al. 2008). Much of engineering practice is directed toward risk management, which has previously revolved around fixed specifications. For example, a building code may specify that the designer shall use a factor of safety of 1.7 in designing against live loads (ASCE 2013), which makes the engineer's job easier since some group of experts has done this probabilistic analysis work already (de Neufville 2004).
Essentially, engineering does not typically design for a range of possibilities (de Neufville 2004). This is why incorporating and developing a new comprehensive approach to risk management and planning will be so challenging. There are numerous design variables that introduce some level of uncertainty, but SLC has an especially wide range of possibilities from 0.3m to 2.5m by the end of the century (Church et al. 2013;Stephens et al. 2017;Sweet et al. 2017).

Industry efforts address the SLC risks
Federal agencies recognize the need to incorporate climate change factors such as SLC into infrastructure design, but most design changes occur for structures that are being rebuilt after being damaged or completely destroyed (Savonis 2014  With the abundance of information out there, the logical next step is to understand how port and marine infrastructure engineers utilize this information and in what capacity this information is being incorporated into design.

Related research on port stakeholder considerations of climate change impacts
Previous research suggests the importance (Becker et al. 2013;Becker, Toilliez and Mitchell 2015) and the difficulty (Milly et al. 2008;Ahern 2011;Olsen 2015) of designing more resilient infrastructure, but there is little understanding of the current state of the practice. Surveys have targeted port directors and other port operations personnel to gauge climate change planning efforts more generally (Bierling and Lorented 2008;Becker et al. 2012). While port directors play a role in planning for SLC, port engineers often make final determinations about how to incorporate SLC into infrastructure designs. Thus, the survey described in this paper focuses on engineers and their decision making processes to assess how SLC is currently considered in the design of port and marine infrastructure.

Participants
The survey targeted engineers from consulting firms, port authorities, and government agencies who have any experience working on a wide variety of port After distributing the survey, 85 useable responses were received. The majority of all respondents work for private consulting firms (60) while those working for port authorities (16) and government agencies (9) make up the remainder of the respondents. Of the 85 responses, 62 respondents voluntarily provided the name of the organization they work. From these responses, 31 different private consulting firms and 11 different port authorities were represented, providing a broad state of the current practices for designing port infrastructure for SLC across the United States.
There were nine private consulting firms represented which had more than one respondent for the firm; six firms with two respondents, two firms with three respondents, and one firm with four respondents.
Results showed that 59% of respondents had over 15 years of experience, 81% self-identified as a project manager or someone who makes final design decision on projects at their organization, and 68% reported spending more than half of their time working on port infrastructure projects as opposed to working on other, non-port related, infrastructure projects.
Respondents were also asked to indicate their experience across different geographic regions. There were 46 respondents (54%) that indicated having experience in more than one geographic region. The region with the greatest number of respondents was the Gulf Coast (42), followed by Alaska (38)

Procedure
In September 2018, the survey was first distributed to all members of COPRI's Ports and Harbors Committee through SurveyMonkey, a web service for conducting surveys. This approach allowed the researchers to easily reach a population spread across the U.S. The survey was also incorporated into the COPRI October 2018 newsletter, posted to the "Environmental, Coasts, Oceans and Water Infrastructure" forum within ASCE Collaborate, posted to the Coastal List (Center for Applied Coastal Research 2019), and shared on LinkedIn through SLC Subcommittee members' profiles. Furthermore, a snowball sampling (Atkinson and Flint 2001) approach was also employed where recipients of the survey were asked to distribute the survey among their professional networks in order to ensure a robust sample size that represents the nationwide engineering practices of designing port infrastructure for SLC. Through these outlets, researchers believe they reached a representative portion of practicing port infrastructure engineers in the U.S.

Measures
The online survey instrument (Appendix 3A) was developed with a five-person project steering committee consisting of members from COPRI's Sea Level Change Subcommittee. The 20-item survey was designed for practicing engineers in the U.S. and estimated to take 10-15 minutes to complete. The survey was broken down into four sections: general information about the respondent, the capacity in which SLC is being considered in design, how SLC is being incorporated into design, and perceived barriers to incorporating SLC into design. To validate the survey for the intended audience, the survey was pilot tested by other members of the Ports and Harbors Committee (i.e., retired professional engineers, engineering professors, and regulatory engineers).
Nine out of the 20-items are presented in this paper in order to answer the three

Data cleaning
In total, 118 responses to the survey were received, and 85 of the responses were useable. There were 12 respondents that indicated having no professional engineering and/or design experience working on port infrastructure projects, and 21 responses with no questions answered for the final three out of four sections of the survey. Therefore, those 33 responses were excluded. Partially incomplete responses, however, were included in the analysis.
As discussed in Section 3.3, Likert scales were used in four of the nine survey questions addressed in this paper. Due to scarcity of the data for some of the Likert scale response options, the data was bifurcated. For Q8, Q12, and Q13 which used a frequency Likert scale, response options Never and Rarely were grouped, as were Often and Always, thus leaving three groups, including the third as Sometimes.
Similarly, for Q14 which used a level of confidence Likert scale, response options Not at all confident and Little confidence were grouped, as were Fairly confident and Very confident, with Neutral as the third group. For Q10, which asked about whether or not the respondent's organization had a formal document that communicates how SLC should be incorporated into design, responses options were grouped for simplified analysis. Those who responded Yes, and we use it for all projects; Yes, and we use it for some projects; and Yes, but we rarely use it were placed in one group ("Have Policy Document"). Those who responded No or I am not sure were placed in another group ("Don't Have Policy Document or Not Sure").

Results and Discussion
The results and discussion section describes survey respondents' perceptions of the state of the practice for designing port infrastructure for SLC, including organizational policies, sources of scientific data on SLC, SLC implications for design life, and reasons that projects do or do not consider SLC. This section uses the results of the survey to provide evidence with respect to the three research questions.

Capacity in which SLC is incorporated into design specifications
The first broad question the survey intended to answer was: In what capacity are port infrastructure designers incorporating a SLC projection into their design specifications for large-scale port engineering projects? This question aims to identify the current level at which engineers consider SLC and to produce baseline data to track how the state of the practice changes in the future. Appendix 2 provides a list of structure types that this survey considered "large-scale." Respondents were asked the total number of port infrastructure projects they worked on in the past five years and the number of those projects that had incorporated SLC. On average, respondents played a role in designing 11.1 (SD: 9.9) port infrastructure projects in the past five years. Further analysis suggests that on average, 43% (SD: 39%; Median: 30%) of port infrastructure projects that respondents worked on over the past five years have incorporated SLC. Because engineers with more SLC design experience may have been more likely to respond to this survey and skew the results, 43% may not be an accurate nationwide indicator of the capacity in which port infrastructure design incorporates SLC. It is likely that 43% is optimistic due to the potential sample bias.
To better understand the organizational level approaches to designing for SLC, respondents were asked if their organization has a policy or planning document that communicates how future SLC should be incorporated into port infrastructure projects. As shown in Figure 1a, 64% of respondents indicated that their organization did not have a policy or planning document, with only 29% having a document, and of those respondents, 9% use it for all projects, 16% use it for only some projects, and 4% use it rarely. The remaining 7% were unsure whether their organization had a SLC design document.
The responses to this question were then used to assess whether or not the organization having a policy or planning document had an effect on the number of projects that incorporated SLC (Figure 1b). There were 25 respondents (18 of which represented private consulting firms) in the "Have Policy Document" group and 60 in the "Don't Have Policy Document or Not Sure" group. Within each group, the average percent of SLC incorporated projects that respondents worked on in the past five years was calculated. A difference between the two groups was found where the average percentage of projects that have incorporated SLC is 30% higher for respondents that work for an organization with a policy document (Mean: 65%; SD: 35%; Median: 67%) than those who do not (Mean: 35%; SD: 36%; Median: 20%). One explanation of this difference is that engineers who work for an organization with a formal document have been specifically informed on how to approach designing for SLC, which might give them confidence to recommend design changes to a client or an in-house design team. Additionally, document provides solid ground for the engineer to stand on when making these recommendations. Conversely, engineers without the documented support from their organization may be less willing to take the personal and professional risk that comes with making subjective decisions on how to incorporate SLC into a project. If an organization has not formally outlined how to design port infrastructure for SLC, it is potentially less likely that their engineers will incorporate SLC because they do not have the necessary protocols or tools to do so. Without a formalized or even knowledge of a formalized document, practitioners must wade through the differing guidance produced by various agencies.
Having a policy or planning document at the organization level could also become a selling point for the organization in competing with other private consulting firms for a contract. Port authorities are beginning to require SLC considerations in the design and redesign of port infrastructure more frequently. In 2018, for example, the Port Authority of New York and New Jersey (PANYNJ) sent out a request for proposal (RFP) for the replacement of numerous wharf structures, which required the bid to provide best practice wharf design concepts that take into account sea level rise (PANYNJ 2018). As these projects and practices become more prevalent, private consulting firms that have a clear and specified approach to designing for SLC could have an advantage.
Researchers also examined the capacity in which SLC was incorporated across different types of port infrastructure projects. Respondents were asked how frequently their organization considers and/or incorporates SLC into the design of 17 different types of port infrastructure. To make comparisons between structure types that are similar in functionality, the 17 infrastructure types were grouped into six different subgroups: protection structures, berthing structures, cargo storage structures, connectivity infrastructure, electoral and operations, and water flow structures.
Appendix 2 contains the descriptions of each structure type. Figure 2 shows how often respondents believed their organization incorporates SLC for each structure type. The researchers grouped responses into three frequency categories. The y-axis shows the percentages of responses for each category. The structure type is on the x-axis, and the number of respondents with design experience for each structure type (n) is indicated in parenthesis next to or below the structure type. and Vessel Berthing Structures (dock structures and wharf structures). Conversely, port hinterland connections such as roads and railways, which are typically located further away from the waterfront, had two of the four highest percentages for either rarely or never incorporating SLC. Understandably, these findings suggest that the closer to the water a structure is located, the more likely the design of that structure will incorporate SLC.

Incentives and disincentives for incorporating SLC into design
The second research question inquired about: where do incentives and disincentives originate for U.S. engineering firms to incorporate SLC into the design specifications of large-scale port engineering projects? First, this section discusses the variety of factors that can act as an incentive to incorporate SLC and how the decision can originate from several different stakeholders. Conversely, for projects that do not incorporate SLC, this section then addresses the disincentives that prevent engineers from incorporating SLC and how the development of regulatory design standards can alleviate several barriers identified by respondents.

Incorporating SLC into design is motivated by a variety of factors
To better understand the driving forces behind designing port infrastructure for SLC, respondents were asked what factors cause their organization to add a SLC design component to a project. Realizing that for any given project, the decision to incorporate SLC could originate from any, or even a combination of these factors, respondents were asked to indicate how often each factor plays a role in causing SLC to be incorporated into a project (Figure 3). Figure 3 -Potential factors that may cause port infrastructure engineers to incorporate SLC into the design specifications of port infrastructure projects. There was no standout factor that always drives the incorporation of SLC into design Client requirements, engineering recommendations, and regulation requirements were the three leading factors that respondents suggest drive the incorporation of SLC in port infrastructure. However, the Often/Always group had the greatest percentage of respondents for four of the five factors. Incorporating SLC based on a life cycle cost/benefit analysis was the only exception. In this case, Never/Rarely (35%) was the most common response, but only by 1% over the Often/Always group (34%).
Simplifying the five factors listed, one factor is client dependent (Client requirement), one factor is regulatory dependent (Regulation requirement), and the other three factors are in the hands of the engineers (Engineer makes recommendation to the client, Design alternative presented to the client, and SLC is incorporated based on life cycle cost/benefit analysis). Responses to this question suggest that none of the three groups are leading the effort to incorporate SLC. Furthermore, responses suggest the decision to incorporate SLC could originate from different stakeholders from project to project.
Although there were only slight variations in the responses, and there were no factors that stood out as being the least likely driver of SLC consideration, Incorporating SLC based on a life cycle cost/benefit analysis had the greatest percentage of respondents that said it was either Never or Rarely a driving factor.
Perhaps engineers are not conducting a life cycle cost/benefit analysis, which would indicate a lack finances or incentives to execute long term planning, or engineers are conducting a life cycle cost/benefit analysis, but the results of the analysis suggest it would be more cost effective to ignore SLC. Further investigation into the use of life cycle cost/benefit analysis, long term planning from the engineering perspective, and design life challenges would shed more light on why this particular factor appears to play a very limited role in the decision to incorporate SLC into port infrastructure design. Additionally, although regulation was only the third most common factor, it is possible that regulation may be the leading factor in some geographic regions, such as California ("State of California Sea-Level Rise Guidance" 2018), but a non-existent factor in other regions, such as the Gulf Coast.
To gauge how engineers determine the level of SLC they need to design for, respondents were asked from where their organization obtains SLC projections ( Figure 4). As mentioned in Section 2.2, numerous organizations have produced SLC projections with varying rates. Therefore, engineers must make decisions on which projections they will rely on. According to respondents, the most frequently used source of SLC projection data was NOAA (65%), followed by USACE (49%).
Although the third most commonly used source was state or local organizations (40%), there were an equal percentage of respondents who rarely or never use state or local projections (40%). For five out of the seven sources shown in Figure 4, at least half of the respondents indicated either sometimes, often, or always using that particular source. Therefore, outside of the fact that NOAA and USACE are the most relied upon sources, these findings highlight that there is very little, if any, standardization across the approach taken by different organizations to incorporate SLC into design. It appears that any one organization could use different sources of SLC projections for any particular project. This could be due to different clients requesting the use of different SLC data, but it could also suggest that there is a lack of consistency across planning for SLC.  Of the 70 respondents to this question, 36 indicated that having no standards was a reason for not incorporating SLC. A lack of project funding and the client not wanting to incorporate SLC were tied as the second most commonly acknowledged barriers. Furthermore, 17 of the respondents indicated that there were other barriers that they felt prevented SLC from making it into final design which were not included the list provided within the survey. Other barriers included site constraints, raising certain structures for future conditions renders them unusable during current tidal conditions, and difficulty incorporating SLC for retrofit, rehab, and upgrade projects on structures that were not originally designed for SLC. Figure 6 shows that every barrier listed was seen as a potential barrier by at least 10 respondents. These findings suggest engineers felt that numerous barriers prevent SLC from being incorporated into design. Figure 6 -Potential reasons why SLC may not be incorporated into the design of port infrastructure projects. Responses were ranked from left to right in descending order of the total number of respondents who confirmed that the potential reason provided in the survey was in fact a barrier that can prevent SLC from being incorporated into design The only barrier that more than half of respondents (36) acknowledged was No standards. Regulatory standards and codes remove the burden on engineers to make their own subjective decisions when determining how to incorporate SLC and make other barriers less relevant. For example, design standards would override the client's decision, which respondents saw as both a major barrier in SLC design and as a primary driving factor in the decision to design for SLC. The results show that the client clearly has a majority of the decision making power, but standards provide consistency in SLC design specifications. Additionally, Lack of project funding would no longer hinder the incorporation of SLC. One respondent noted, "Lack of planning or vision for surrounding facilities being modified for sea level change has caused accommodating for sea level change to be the first item removed from scope of project to meet funding." When funding is limited, SLC can be low on the priority list.
However, as another respondent alluded, removing SLC from the scope of a project would not be an option if there are regulatory design standards in place, stating, "The cost differential cannot be justified, especially when it is not a regulatory compliance issue." As mentioned in Section 2.3, federal regulation has had some success under Executive Order 13690. Until it was revoked, EO 13690 provided clear flood protection requirements for designing infrastructure. Although SLC is projected to be highly varied across coastal regions of the U.S., these requirments provided flexibility and allowed owners and engineers to select one of three options: use best available climate science, design for the 1-percent-annual-change flood (100-year flood) plus an additional two feet, or design for the 0.2-percent-annual-chance flood (500-year flood). This flexibility alleviates some of the finacial stress by not forcing a specific action onto an owner. Of course, some ports have a greater institutional capacity to cope with these requriments, but providing different options minimizes any strategic advantage one port would have over another when requiring all U.S. ports to address increased flood risk.
Restablishing EO 13690 would be a positive approach toward improving the resilience of seaports nationwide. However, as discussed in Section 2.2, the concept of stationarity and utilizing 100-year flood or 500-year flood events to guide design is an outdated one. Due to climatic changes, the return period probability for storm events is no longer what it once was. The entire globe is witnessing more intense storm events and more frequent high intensity storms, so it can no longer be accurately predicted what a 100-year storm brings in terms of flood extent and depth.
ASCE has a unique role to play in the development and improvement of regulatory standards and codes. ASCE has proven to be a leader in the develpoment of flood resistant design standards though ASCE/SEI 24-14 (ASCE 2014). ASCE 24 is the industry standard for flood-resistant design and construction, and has been adopted by several building codes (Ayyub 2019). However, ASCE 24 does not adequately address the implications that SLC can have on design flood elevations (ASCE 2014).
An updated verion of ASCE 24 that does account for SLC would be a significant benefit toward implementing regulatory desing standards across the nation. Just as EO 13690 was developed with input from the engineering community, any future federal regulation should also be crafted in a collaborative setting. In Canada, the engineering profession believes that engineering codes, standards, and work practices should consider climate change, and that federal and provincial governments must collaborate with the engineering profession on climate change policies for the benefit of the public (Engineers Canada 2013).
Alternatively to federal regulations, states could take it upon themselves to increase the resileince of their port infrastructure. In the context of seaports, state transportation agencies could establish and enforce regulatory design standards. Some states have already deemed their current codes and standards inadequate to prepare infrastructure for a changing climate and wrote their own standards (Ayyub 2019). For example, in California, Executive Order S-13-08 was issued in 2008 which directed state agencies to plan for SLC. This led the California Department of Transportation to develop guidance on incorporating SLC into transportation infrastructure projects (Caltrans 2011).

SLC uncertainty challenges are minimized by avoiding long-term design decisions
The final research question asked: For engineering firms that are incorporating SLC, what strategies are the port infrastructure designers in those firms implementing in the design specifications of large-scale port engineering projects to cope with the scientific uncertainty of SLC? Some respondents indicated designing port infrastructure in a way that can accommodate future upgrades to keep pace with SLC (Results Appendix 4D), but results ultimately suggest that SLC uncertainty has not been a major consideration due to relatively short design lives, where uncertainty is not as significant as it is toward the end of the century. Therefore, strategies to cope with uncertainty have not been widely developed or implemented. Service Life -The period of time in which the structure is actually in use, from construction completion to failure.
prohibitively huge to raise marine deck structures." Another wrote, "So much of the work is retrofit of existing docks that generally it doesn't make financial sense to raise." And a third reported, "It is hard to accommodate significant sea level rise with existing large marine terminals (multiple thousand feet of wharf, 200+ acres, rail, etc.)it is not financially feasible." This raises attention on how new infrastructure is designed, and highlights the importance of new infrastructure incorporating SLC.
Furthermore, these findings call into question port planning time frames and the rigid methodology of designing structures for a specified lifetime or "design life" rather than the structure's "service life" (Figure 8).
Design life of port infrastructure varies depending on structure type, but typically ranges from 30-50 years. However, it is not uncommon for some port infrastructure to have service lives that exceed 100 years (Becker, Toilliez and Mitchell 2015). This is a concerning disconnect when designing port infrastructure for SLC. For example, new infrastructure designed for the projected sea levels of 2050 could likely remain in service beyond 2050. Therefore, the design may be inadequate for the change in sea level between 2050 and the end of its service life. Alternatively, the structure could be repaired, retrofit, or upgraded at the end of its design life, but as  (Biondini and Frangopol 2017). A study conducted by ICF International, Inc. also highlights the benefits of this approach where the authors support LCCA by saying, "This methodology can be used to support decision making regarding climate change adaptation alternatives under compounded uncertainty. In addition, this methodology can be used to determine which adaptation design alternative is the most consistently resilient across the range of climate change and disaster event scenarios" (Rodehorst et al. 2018). This could be a potential way forward for engineers of port and marine infrastructure to navigate the challenges of designing for SLC.
The results of this survey also present new opportunities for future research that addresses climate change adaptation and resilience for seaports. The inconsistencies in approaching SLC design challenges and the lack of SLC design standards highlighted in this paper calls for collaboration among the engineering community, port authorities, and regulating bodies to improve the resilience of port infrastructure. Developing design standards collaboratively can help engineers overcome the barriers that currently prevent them from incorporating SLC. With design standards in place, many of the other barriers acknowledged by respondents would no longer exist. Therefore, further exploration and discussion is required to determine the most effective approach to implementing design standards. Should regulation be implemented at the federal level? Should states be the ones to develop their own design standards? Should standards be applied based on design life? Should standards be specific to the type of infrastructure? These questions deserve further dialogue as SLC becomes an increasing threat to port infrastructure. Also, the opportunity exists for private consulting firms that have a policy or planning document for SLC design to share resources, tools, and best practices with other members of the engineering community. Organizations such as ASCE have proven to be great facilitators of this type of knowledge sharing. Given the massive amounts of infrastructure spending that ports are planning in the next five years, ensuring that these investments are sustainable and resilient to future environmental conditions should be a top priority.

Limitations of Research
This was the first nationwide survey of port and marine infrastructure engineers regarding the practice of designing port infrastructure that is resilient to SLC. The sample originated with members of COPRI who have port infrastructure design experience, and expanded through snowball sampling. There were at least 31 different private consulting firms and 11 different port authorities represented in this sample, but there are clearly port infrastructure engineers at other consulting firms and port authorities across the country. It is difficult to determine the total number of consulting firms in the U.S. that work on port infrastructure projects, and therefore, it is difficult to gauge what portion of the entire population responded to the survey.
Survey recipients who did not respond may not be interested in SLC design issues. Therefore, responses may be skewed toward engineers who are aware of the challenges brought by SLC and who have more experience designing port infrastructure projects for SLC.
The researchers designed the survey to gauge the general state of the practice across the U.S. Therefore, the results are not indicative of engineering practice within specific regions, and are not indicative of engineering practice outside of the U.S. SLC impacts will vary, resulting in SLC design challenges to become a greater priority in some regions. The survey was not designed to identify the location of port infrastructure projects that respondents have worked on. Since 46 respondents acknowledged having experience in multiple different geographic regions, results cannot determine the location of each project a respondent worked on, therefore, responses could not be analyzed separately based on geographic location.
A sample of engineers within specified geographic regions would provide findings for further comparison. This survey was designed for engineers working for private consulting firms, port authorities, and government agencies. As a result, the researchers may have overlooked potential differences in SLC design approaches between these groups. Separate surveys that target each group individually could reveal differences in approach to designing for SLC. While additional details of engineering practices need to be explored in this area, the researchers feel that this study provides informative baseline data where key issues in the resilient design of port infrastructure can be identified and addressed.
Bifurcated Likert scales used for presenting results is also a limitation of the survey and data collected. Although data analysis was conducted with and without bifurcating the data and results were similar in each instance, results had only slight variations when the data was bifurcated. However, due to the similarity in results of the non-bifurcated data and the data that was bifurcated as discussed in Section 3.3, the researchers believe that bifurcating the data in this way was acceptable.
Potentially gathering a more international dataset or even tailoring the survey to specific audiances or stakeholders, this could provide more future variability in the outcomes.

Conclusion
SLC design decisions made today have long-term impacts on the resilience of port infrastructure. Engineers must consider SLC and climate change impacts when designing port infrastructure to ensure that ports can continue to serve their essential role in the global economy in the coming decades. In serving the public interest, engineers are uniquely qualified and positioned to ensure port infrastructure is resilient for future sea level scenarios. However, adequately designing port infrastructure for SLC is a challenging task due to the uncertainty of SLC projections and the long service lives of port infrastructure. The inconsistencies revealed by this study suggest that the incentive to incorporate SLC into design is inconsistent from project to project, as are the barriers that prevent incorporating SLC into design. Furthermore, SLC projection data varies across NOAA, USACE, IPCC, state and local organizations.
To overcome these challenges, the engineering community needs to develop systematic and practical methods for incorporating SLC into design decisions.
Engineers have the opportunity to work with both port authorities and regulatory bodies to help improve resilience efforts through knowledge sharing of successful design strategies, the development of design standards, and transitioning away from the traditional frameworks that engineers have operated within. Only 29% or respondents indicated that their organization had an internal policy or planning document that communicates how to design for SLC. Knowledge sharing between organizations could promote the adoption of formal guidelines and lead to more consistency in the engineering community's approach. Findings also suggest that the lack of design standards in this area leads to engineers and their clients disregarding SLC more frequently. Regulatory design standards would alleviate some challenges, but the development of such standards needs more robust dialogue between engineers and regulators. Lastly, designing port infrastructure for a theoretical point in time can leave structures at risk to SLC beyond its design life. Future retrofit and upgrade projects can ensure structures are resilient throughout their service lives, but as respondents noted, it is much more difficult to incorporate SLC as an afterthought.
Engineers can help improve the resilience of port infrastructure by transitioning to a life cycle cost analysis design approach, which is better suited to address the uncertainties that climate change introduces.
The appendixes below describe the aspects of this study in greater detail. They provide greater detail on different types of port infrastructure, life cycle cost analysis design, additional methods, and additional results obtained from the survey. We also provide a copy of the survey instrument.

Appendix 2: Defining Port Infrastructure
This appendix provides more information on what the researchers considered "large-scale" port infrastructure projects, as well as the different types of port infrastructure that were listed in question eight of the survey: For which types of structures does your organization incorporate/consider sea level change during the design phase? Table 1 provides definitions for each structure type. In the survey, five questions required respondents to answer on a 1 -5 scale, which included (1) Never, (2) Rarely, (3) Sometimes, (4) Often, and (5)  Responses to these questions were analyzed in two ways. First, the researchers treated the five primary response options separately in the analysis. Secondly, from the five primary response options, three groups were created: One group for Never and Responses to these questions were also analyzed in two ways. First, the researchers treated the five primary response options separately, and then placed responses into three groups: One group for Not at all confident and Little confidence, a second group for Neutral, and a third group for Fairly confident and Very confident.
After comparing the two analyses, no clear difference in the analysis emerged.
Therefore, results were presented using only the three groups.