This chapter identifies various applications of urban meteorological products and services of value to a range of end users, as informed by the end users who attended the workshop. There is little peer reviewed literature on this topic and much of the available information is anecdotal in nature. Given this, the Committee does not offer conclusions, but rather reflects on the workshop discussions and highlights key themes that emerged.
End users of urban meteorology information, such as urban planners, emergency planners, and local utilities in both the public and private sectors of society, have demonstrated important needs for urban meteorological observations and analyses. Policy makers, managers, and regulators at the local, state, and federal levels, as well as those in sectors such as public health, emergency response, security, electric power, flood control, transportation management, urban design, business, and the general public, use urban meteorological information—both real time weather and long-term climate information—either directly or indirectly (e.g., through use of outputs of environmental health and science work that applies urban meteorological data) for planning and decision-making (Box 2.1; also see Dabberdt abstract in Appendix A). There are likely several additional end user groups that have not been identified in this report. The workshop was an attempt to reach out to some of these groups and a few papers and reports have also described end users and their needs (e.g., Dabberdt et al., 2000; OFCM, 2004; Grimmond et al., 2010b; Mills et al., 2010). However, an important question remains unanswered: How can the urban meteorological community better identify these groups, reach out to them, and begin an ongoing dialogue to assess and better meet their needs?
Examples of How Some End Users Apply
Urban Meteorological Information
Urban meteorological observation and modeling results are used in various applications by a variety of users. The types of data, as well as the temporal and spatial resolution required by these end users, vary substantially by application (Dabberdt abstract, Appendix A).
Decision makers in the insurance sector may use monthly climate summaries to validate industry catastrophe models, which simulate the geographical risk associated with natural and man-made catastrophes.
Retailers, such as local grocery stores, use short-term weather information, particularly about winter weather events. Managers of large-event venues (indoor and outdoor), such as sporting events, fairs/carnivals, concerts, etc., also use short-term weather information to make decisions about canceling events. Larger retailers, such as clothing stores, use longer-range weather and seasonal climate information.
Urban planners use meteorological information to assist in the development of climate change mitigation and adaption strategies of cities and regions. Urban planners could also use weather and climate information to help select urban locations that can ameliorate weather-related human suffering and mortality. Indeed, this would be a good subject for future multidisciplinary studies.
Environmental protection agencies at federal, state, and local levels use meteorological information to monitor, regulate, and set standards for the protection of human health and the environment.
County officials, educators, and school superintendants use meteorological information to make decisions about closing schools during winter weather and convective weather events.
Personal Decision Support
The general public makes personal decisions (e.g., weekend activities, outdoor recreation, wedding planning) in daily life.
Basic and applied researchers use meteorological information to conduct research in the various fields discussed in this report.
End users of urban meteorological information are heterogeneous and cover a vast spectrum of job roles, goals, needs, and understanding. Moreover, different end users may be more or less “advanced” in their use of meteorological data for a host of reasons. For example, some users have been working with the meteorological community and using weather data for a longer period of time. Some users have more resources (people, time, money, etc.). Some users have different knowledge (educational or experiential) bases. These differences can vary within and among groups and geographic areas. A given end user’s needs may also vary depending on the type of weather observation or phenomenon being considered. Acknowledging and understanding this heterogeneity is important for the urban meteorological community to better understand, interact with, and meet the needs of end users.
Furthermore, end users may be viewed as a “cascade” or “web” of individuals and groups with varying information needs and at varying levels of distance from raw urban meteorological data (Box 2.2). Given this interconnected relationship, it is important to recognize that there are multiple types of urban meteorological phenomena that have impacts on different types of users with different types of needs (see discussion in Dabberdt Abstract, Appendix A). Accurate information on low-risk, high-frequency events (e.g., regular weather patterns) may be of importance to a broad range of end users; whereas knowledge of high risk, low frequency events (e.g., a chemical release emergency) are of utmost importance to emergency planners responsible for public safety in such circumstances. To complicate long-term planning, both types of events will be affected by climate change. As such, end user needs span a wide range of urban meteorological information, from high frequency and low frequency events that may occur over the short and the long term.
As many participants noted at the workshop, end user needs with regard to urban meteorological information currently are not sufficiently being met (Table 2.1). In many cases, urban meteorologists are simply unaware of the precise data and information needs of the various information groups. However, there is a risk that if urban meteorologists do not provide the required information, disparate groups of end users will start generating (or will more fully develop) their own data streams, not necessarily following best practices in data collection, analysis, or interpretation, and producing not only redundant but also inconsistent information. More importantly, if the urban meteorological community does not provide required information, end users’ needs will not be met, reducing the effectiveness of their decision-making.
Cascade of Users
There are many different end users that require some form of urban meteorological information to help them make informed decisions in their respective fields. End users have a vast array of data needs and require this information at different spatial and time scales. Some end users work with raw data directly, some share the data they have acquired with other users (e.g., a water resource public utility may share their rainfall projections with the emergency response community so they may alert the public), and many work with the research community and weather services to garner the data and information they need. Some end users also share their own urban meteorology data with the research and weather services communities. The general public, itself an end user, often makes decisions based on information from the weather services community or through information provided by other end users. This “cascade of users” diagram is complex and meant to be illustrative (see figure on next page). It does not depict every user of urban meteorological information nor does it depict every pathway that this information is shared among users. Directions of arrows denote the flow of urban meteorological information. Raw data includes both in situ and remote sensing observations. The end goal of the information sharing throughout this cascade is better informed decisions by decision makers.
It is also important to note that the end user needs and data needs highlighted in this report are not necessarily equivalent. Data typically needs to be translated into useful information and/or decision support products for end users to utilize. The following sections consider some of the major unmet needs identified by participants at the workshop who use meteorological information in their regular decision making, followed by a discussion of communication needs across disciplines.
Over the last few decades, clear advances in urban monitoring and forecasting have been made (described in detail in Chapter 3). The needs discussed below were identified at the workshop by end users as being both “significant” and “unmet.”
Targeting Observations and Models for User Needs and Preferences
One key role of science in society is to provide information that can be applied to real problems (i.e., tailored, information-rich products and
services that decision makers can use effectively). Figure 2.1 is an example of a heat vulnerability framework advanced by Wilhelmi and Hayden (2010) that illustrates such connectivity.
Over the last decade, the emergence of urban forecasting decision support systems for end users—taking into account user needs and preferences
TABLE 2.1 Sampling of specific unmet end user data needs
|Sector||Examples of Unmet Data Needs|
(municipal and public safety officials)
|• Rainfall and snowmelt runoff and storm water datasets
• Urban flooding and/or overloading of combined storm water/sewage systems due to localized precipitation and ability/inability of urban pervious surfaces to store water
• Atmospheric river (i.e., narrow corridors of concentrated moisture in the atmosphere that when striking land can produce hazardous storms) information
(power producers, grid operators, local utilities)
|• Air temperature for assessing energy demands and related loads on the grid
• Wind and solar radiation data for renewable energy assessments
|• Accurate and timely forecasting of extreme events
• Surface roughness, overland decay, and wind speed
(company officials, public and private service providers)
|• Solar radiation, precipitation, and air quality data for agriculture (e.g. for agricultural regions near and/or impacted by cities)
• Canyon-level wind flow (e.g. for construction sector)
(architects, urban planners, municipal officials)
|• Vegetations stress index for cities/optimization
• Urban air quality
• Assessment of urban heat island mitigation measures such as green roofs and tree planting campaigns
• Development of climate change mitigation and adaptation strategies of cities and regions,
• More dense array of first order meteorological stations in and around urban areas
• Improved methods for assessing the extent to which rural meteorological stations are subject to the impacts of local land use change
(officials in departments overseeing highways, railroads, airports, harbors, and rivers)
|• Canyon-level wind flow
• Precipitation and its form (i.e., rain, freezing rain, sleet or snow)
• Representativeness of surface observations
• High spatial resolution forecasts (e.g. roadway scale)
• Road surface temperatures
|Sector||Examples of Unmet Data Needs|
(health department officials, environmental protection agency officials, air quality management districts, public safety officials, emergency managers)
|• Solar radiation, wind, humidity and air temperature at matching scales for health (e.g., heat indices)
• Consistent urban heat island baseline datasets for vulnerability/risk assessments (standardized methods and data)
• Spatially explicit datasets that characterize the urban heat island (i.e., further than just surface air temperature measurements; surface skin temperature, air temperature, humidity, wind and radiation data may provide a more comprehensive assessment of “heat”)
• Heat and cold wave and physical stress forecasts with temporal and spatial resolution at city scale
• Street-level air quality
• Extreme precipitation event forecasts
• Extreme localized heat/cold advisories, disease vector, and air quality advisories
(public safety and security officials)
|• Higher temporal, vertical, and horizontal spatial resolution data (e.g. urban boundary layer structure and mixing layer heights, vertical profiles of winds, turbulence, temperature of particular importance to dispersion applications)
• Dual-use leveraging of data from other applications (e.g. radar-derived precipitation calibrated with rain-gauge data for flood predictions)
• Regularly updated urban data (e.g. land-use characteristics, building footprint data)
(public and industrial safety officials)
|• Street-level detailed flood information
• High spatial and temporal resolution wind, temperature, and moisture data in and above the urban canopy
from the very start—is a significant advance. Increased data interoperability and coupling of physical and social science data and models on various scales (global to sub-city scale) provide new opportunities to assess vulnerability, refine decision making, and respond to threats or hazards. Common synthesis of data on exposure will facilitate research on urban vulnerability, impacts thresholds, and adaptive capacity. The outcomes of such integrated research then can be applied via a decision support system. One example of this has been the development of the Maintenance and Decision Support System (MDSS) in 2001 (Box 2.3). Prior to its development, most State Departments of Transportation (DOT) had very little objective information on road conditions. The MDSS provides 72-hour forecasts of atmospheric and weather conditions for many snow-fighting states.
FIGURE 2.1 Heat vulnerability framework (following Wilhelmi and Hayden 2010). SOURCE: Wilhelmi and Hayden, 2010.
Indiana’s Implementation of MDSS
A Maintenance Decision Support System (MDSS) is a tool that utilizes forecasts, predictions, and reports on observed weather and road conditions to assist managers in making informed decisions on how to best utilize resources when planning for and treating snow and ice (see figure below). It also serves as a training tool that can be utilized year round.
In the late 1990s, the Indiana Department of Transportation (INDOT) decided to begin utilizing new technologies to more effectively prepare for and mitigate the impacts from snow and ice. Representatives from INDOT attended one of the Federal Highway Administration’s (FHWA) MDSS stakeholder meetings and saw potential in the developing MDSS project, which was created with input from national labs and private contractors.
INDOT was facing declining revenues and eventually progressed from limited use of an MDSS to statewide implementation during the winter season of 2008-2009, largely because of the savings reported by the maintenance units utilizing a MDSS. Compared to the previous winter season, MDSS helped INDOT achieve savings of over $12 million in salt usage and $1.3 million in overtime compensation by the end of the 2008-2009 snow and ice season (McClellan et al., 2010).
Before statewide implementation, effective training had to be provided because a majority of INDOT staff had no experience with MDSS. An introduction to MDSS was presented at the 2008 INDOT Snow and Ice Conference, and eventually six training packets were supplied to relevant end users to ensure that they had the working knowledge required to make the MDSS project useful. In addition to extensive training, forecasts were fine tuned and recommendations to match actual observed conditions were made by communicating with the MDSS vendor during the snow and ice season (McClellan et al., 2010).
To continue the effective use of MDSS and to ensure greatest exposure to MDSS and its capabilities, INDOT plans to increase the focus on hands-on training for the diverse users of MDSS. To date, a group of 14 State DOTs (including INDOT) have pooled their resources to develop a customized MDSS for their respective agencies (Ye et al., 2009).
Maintenance and Decision Support System (MDSS) interface—developed from the start with end user input—provides precision forecasts and treatment suggestions to assist managers in making informed decisions on how to best utilize resources. SOURCE: Sheldon Drobot, NCAR.
Overall, however, there is an inadequate emphasis on providing the necessary observations and modeling activities for meeting the needs of a broad cross-section of urban users. For example, in the quest to create unbiased global-scale temperature records, meteorological services around the world routinely adjust the meteorological record to exclude the urban heat signal (Karl et al., 1988; Peterson, 2003), and monitoring stations affected by urban development are being relocated to remote rural areas. Operational weather and air quality models in the U.S. are also not capable of modeling cities correctly (Grimmond et al., 2011; see Chapter 3 for detailed discussion). As such, the type of data required to study and serve the needs of cities is mismatched with what is currently routinely monitored. It is important that the complementary functions of global and urban climate records be recognized and that the need for quality-controlled long-term urban observations be addressed.
It is crucial that information, including metadata, be collected, analyzed, quality-controlled, and made available to an array of stakeholders and users in formats, timeframes, and presentations appropriate for the application. Sometimes end users do not utilize urban meteorological information because there is a mismatch between the information that is provided to them and the temporal and spatial scales and lead-time of forecasts they need to make decisions. An emergency responder may have very different space, time, and data latency requirements for wind information during a bioterror event within a city than a meteorologist assessing thunderstorm rainfall potential.
Ultimately, it is important that all the communities involved in urban meteorology (e.g., modelers and stakeholders) describe the types of observations and model input and outputs they would need to help ensure the result is a tool that is useful in helping them meet their goals. For example, Seattle Public Utilities has worked with both the local university research and weather services communities to develop and utilize weather products to provide drinking water, drainage, and wastewater services to the residents of Seattle (see Box 2.4). This type of collaboration works to capture commonalities across communities and to optimize return on observation investments and modeling efforts. In order to have the greatest impact, it is important that each community better understand how these resources might best be applied. The leadership of working across disciplines is a key requirement for success.
Data Access and Data Sharing Needs
As discussed above, there is a cascade of end users of urban meteorological information, and this cascade can also be extended to observations
and model products. Many of the users of data also gather their own data, which would be a useful supplement to data collected by other agencies. Many workshop participants asserted that there are important limitations and considerations related to data access, data availability, and data sharing that need to be addressed by urban meteorologists working in partnership with stakeholders.
Data access and availability was identified as a major crosscutting issue at the workshop. Agencies typically steward environmental data that suit their operational needs. For various reasons (e.g., legal or institutional), these data are often not shared externally. For example, government lidar (light detection and ranging) data and National Lightning Detection Network (NLDN) data are often difficult to access but would be a significant source of information for several user communities. Information about the environment (including observations, models and all downstream products) is ultimately a commodity that flows according to forces, restraints, filters, and feedbacks that are applied along the information flow path. These mechanisms can transform the information to generate positive as well as negative outcomes for end users (e.g., economic gain, governmental control, private control, quality of life for urban dwellers, etc.). Recognizing the interplay between these mechanisms and how they affect the outcomes is key for understanding how and why urban meteorological monitoring and forecasting capabilities are not being effectively utilized.
Certain restraints and filters that affect the flow of data are worth noting. First, considerable resources (funds, manpower, infrastructure) are required to render data into a form that is user friendly. For example, most emergency management agencies do not have the resources to translate large amounts of unprocessed data into summary information that informs decision making; with stronger relationships and more collaboration, this challenge could be overcome. Second, there are many throttles that restrain the flow of information, such as data security, proprietary concerns over collected data, researchers wanting to publish their data before providing it to others, and preventing harmful or “bad” information from being projected.
Therefore, in addition to the basic need for data to be made more available to end user groups, it is critical that several data sharing issues, including quality assurance and metadata (Grimmond abstract, Appendix A), are addressed by the research and weather services communities as well as end users. Quality assurance is important because it results in internally and externally consistent standards of the data, which would be useful for effective translation of data from measurement to use. Metadata are especially important because they are all the pieces of information necessary for data to be independently understood by users, to ensure proper
Weather-Related Tools Utilized by the Seattle Public Utilities
(adapted from workshop presentation
by James Rufo Hill, Seattle Public Utilities)
Seattle Public Utilities (SPU) is a department within the City of Seattle that provides drinking water, drainage and wastewater services, and solid waste services to 1.3 million customers (SPU, 2011). Seattle has a misleading reputation for persistent precipitation; the frequency and amount of participation can be quite variable.
Because of this variability, SPU has a long history of managing resources according to the water cycle and is a long-time user and shaper of various weather products. In 2006, SPU partnered with the Mesoscale Analysis and Forecasting Groupa at the University of Washington to design Seattle RainWatch, a forecasting tool that predicts local rainfall patterns for the next hour and provides 1- to 48-hour rain accumulation totals. It uses rainfall estimates derived from NWS radar data that are calibrated with local rain gauges (in particular SPU gauges) to improve accuracy over other radar-only indicated precipitation estimate products. Seattle RainWatch gives SPU a one-hour window to identify which neighborhoods in the city will experience the highest rates of rainfall. Key operators, managers, and crews receive maps generated from Seattle RainWatch to help them quickly identify where resources should be deployed to ensure storm drains are clear and citizens are alerted (see figure below).
SPU also uses SNOwpack TELemetry (SNOTEL), an automated system that collects snowpack and related climate data in the Western United States, including Alaska. Data on snowpack is critical to SPU because 50-80 percent of the water supply in this region arrives as snow (NWCC, 2009).
Streamflow Forecasting Computer Model System (SEAFM), another tool used by SPU, is a model that is designed to repeatedly simulate the current hydrologic state of the watershed. It can also be used to analyze and assess various future reservoir operating scenarios. It produces probabilistic streamflow forecasts up to 12 months by utilizing the latest climate forecasts from the National Weather Service (NWS) (SPU, 2006).
Seattle RainWatch, SNOTEL, and SEAFM are just some of the tools that allow SPU to make short and longer term preparations in managing reservoir levels, ensuring drainage systems are functioning in the city, and providing drinking water.
Example of a 1-hour RainWatch forecast (total precipitation expected in inches). SOURCE: James Rufo Hill, Seattle Public Utilities, http://www.atmos.washington.edu/SPU/
stewardship of the data, and to allow for future discovery (NRC, 2007b). This information should include, at a minimum: a thorough description of each data set—including its spatial and temporal resolution; the time and location of each measurement, and how the data were originally collected or produced—and a thorough documentation of how the data have been managed and processed (NRC, 2007b).
Short-Term Weather: Longer Forecast
Lead Times, Accuracy, and Confidence
To safeguard their services, agencies rely heavily on meteorological forecasts, as do emergency planners for extreme events (e.g., blizzards, heat and cold waves, coastal storms; Figure 2.2). In the transportation sector, for example, each year nearly 700,000 people are injured and over 7,100 perish in weather-related road crashes (FHWA, 2011). Weather is also the leading cause of nonrecurrent traffic congestion, which in turn results in spikes in greenhouse gas emissions (FHWA, 2011). Precise and timely weather information and forecasts enable the traveling public to arrive at their destinations safely and efficiently. A national survey indicates that, outside road closures, information on weather conditions and forecasts are
FIGURE 2.2 Emergency planners typically go through several decision making steps as hazardous weather event information becomes available. The vertical lines in the green part of the figure indicate significant decision points that separate major weather-related decision-making stages. As shown in the blue part of the figure (standard phases of emergency management [Godschalk 1991]), the increased readiness, event specific preparation, and emergency operations stages coincide with the response phase. SOURCE: Morss and Ralph, 2007. ©American Meteorological Society. Reprinted with permission.
the most in-demand items for drivers, more so than information on traffic accidents and points of interest (AMS, 2011). Similar to the traveling public, freight haulers by and large operate with very limited weather information. This leads to the loss of several billion dollars annually (Drobot, 2011). In addition, many emergency medical service (EMS) drivers lack accurate and timely weather information, which can lead to increased risk of mortality in patients (Drobot, 2011).
As many agency representatives at the workshop pointed out, being ill-prepared for an adverse event typically has a much higher penalty than preparing for what turns out to be a non-adverse event, and thus agencies would rather err on the side of caution. Improvement in lead times and forecast accuracy are important because they could help reduce false negatives (forecasting a non-adverse event that turns out to be adverse). This could also help reduce false positives (forecasting an adverse event that turns out to be non-adverse) which could help lessen the unnecessary expense of mobilizing the workforce.
Ultimately, providing better evaluations of the plethora of simple to complex urban models and parameterizations under a wider range of conditions (both weather and land surface characteristics) could help the general public be better prepared for adverse events in urban areas.
Long-Term Prediction: Better Forecasting
of Extreme Events and Trend Detection
Long-term climate change is expected to affect short-term weather patterns, including a potential increase in extreme weather events (Meehl et al., 2000). Urban areas may be particularly vulnerable. For example, the urban environment will exacerbate climate change-related problems like the potential increases in the frequency and intensity of flash floods (with forced run-off, drainage, and traffic problems likely exacerbating the impact) and increases in the frequency and severity of high levels of heat (exacerbated by neighborhood-scale areas of greatly enhanced excess heating and lack of ventilation; see Box 2.5).
Many workshop participants asserted that it is crucial that the field of long-term forecasting, which traditionally has excluded the urban signal, further develop methodologies to include the urban area. Advances in local climate downscaling techniques1 will help with achieving this goal in the medium term. It is important to note that improvements in global climate
1 Downscaling global circulation models to the regional and local level.
Central Texas Climate Change Environmental
Public Health Indicators Tracking Tool
(adapted from workshop presentation by George Luber, CDC)
Austin, Texas, like several cities in the Unites States, is considering climate mitigation and adaptation policies that aim to reduce the threat to public health and local infrastructure. In particular, public health officials in Austin are concerned of the threat from extreme events that may increase due to climate change.
With a population close to 800,000 people, Austin is the fourth-largest city in Texas and the 15th most populous city in the U.S. In order to help make appropriate public health adaptations, the City of Austin, in partnership with the Centers for Disease Control (CDC), developed the Central Texas Climate Change Environmental Public Health Indicators Tracking Tool (EPHI Tracking Tool).
This vulnerability assessment tool aggregates health, weather, demographic, policy, and environmental data and indicators in a Geographic Information System (GIS) viewer at the census block group scale. The goal for this tool is to create baseline indicators of vulnerability to extreme heat and river flooding (identified as the top two extreme events in Travis County, where Austin is located) at which the City’s current and future climate change mitigation and adaptation policy-making priorities can be targeted to reduce burdens on vulnerable populations. The maps generated by the EPHI Tracking Tool increase the City of Austin’s ability to create locally appropriate, targeted interventions, such as identifying areas for tree-planting programs and considering future land-use policies.
In the future, the City of Austin would like to improve this tool by gathering input from stakeholders to enhance the user experience. They would also like to add indicators for air quality and extreme weather evacuees (e.g., impact of increased numbers of hurricane evacuee populations being displaced from the Texas Gulf Coast to Travis County). Tracking tools, such as the EPHI that map health data with environmental data, allow city officials to assess community health vulnerability related to climate change, guide policies, and adapt interventions that are specific and operational at the local level.
models (GCMs) to better capture changes in extremes will also be crucial for improving downscaling techniques.
Trend detection is another important need identified by several users at the workshop. When is the change in climate large enough to implement adaptation change, particularly under extreme events? For example, were the spring 2011 floods in the Midwest the old 500-yr flood, or the new 50-yr flood? Accurately answering questions such as these requires that scientists have a better understanding of the current climate regime. Cities are planning staged adaptation strategies, some of which take a long time (e.g., growing trees) and some that might allow for more rapid impacts (e.g., reflective roofs and pavements). It is critical to work closely with end users to develop methodologies and protocols to identify weather and climate
trends. This work involves both a monitoring aspect (i.e., an understanding of what has occurred retrospectively) as well as extrapolation into the future.
Understanding the Impact of Specific
Environmental Intervention Scenarios
Many cities are deploying large scale greening or albedo modification plans, and models are needed to help justify or understand the expected benefits (or unintended consequences) of these programs (Stone abstract, Appendix A). For example, “vertical gardens,” originally an experiment in 1988 by Patrick Blanc, a French botanist, are becoming increasingly popular in some cities (Figure 2.3).
FIGURE 2.3 A “vertical garden” on Jean Nouvel’s Musée du quai Branly in Paris, France. SOURCE Patrick Blanc, http://www.verticalgardenpatrickblanc.com/.
The Phoenix Urban Form Project
With a population close to 1.5 million, Phoenix, Arizona has experienced rapid suburban growth since the late 1940s. Downtown Phoenix, located at the center of the Phoenix metropolitan region, covers an area of 2.6 square miles. Discussion sparked by a new light rail line connecting downtown Mesa, Tempe, and Phoenix and the decision to locate a new campus for Arizona State University in downtown Phoenix has centered on future development of Downtown Phoenix and how to best utilize the many parcels of empty land surrounding the urban core.
The city initiated the Downtown Phoenix Urban Form Project in order to transform the empty land into high density, mixed-use districts. This new development plan is called the “Connected Oasis” (see figure below). It outlines an “open space pedestrian-intensive network of shaded streets, pocket parks and dedicated ‘green connectors’ tying together major downtown destinations and retail zones” (Studio Ma, 2011). The extreme summer heat in Phoenix poses significant challenges to the design of public open space and influences the new development plan. To balance the competing needs of thermal comfort and the urban heat island effect, the proposed zoning code calls for (1) the creation of a building/street profile that optimizes the flow of air through the building mass to flush out accumulated heat and pollutants; (2) shading of ground and building surfaces to improve the thermal comfort of pedestrians; (3) utilization of highly reflective and emissive building materials; and (4) the creation of “cool pockets” along sidewalks and other pedestrian routes to enhance thermal comfort (Studio Ma, 2011).
Another example of greening, but on a larger scale, is the Downtown Phoenix Urban Form Project, which describes a greening plan for downtown Phoenix. This effort proposes a zoning code that “optimizes building massing and street canyon sections to balance the competing needs of thermal comfort and the urban heat island effect” (Box 2.6). As another example, MillionTreesNYC—launched by the Parks Department and New York Restoration Project, in collaboration with community-based and non-profit groups; city, state, and federal agencies; corporations and small businesses; developers, architects, and landscape architects; private-property owners, and the general public—plans to plant and care for one million new trees in New York City over the next decade.2 Models are needed to characterize how much these efforts will reduce temperatures (and also to explore impact on air quality at street levels, since vegetation may limit ventilation by the wind).
Several workshop participants also identified a need for more scenarios to be modeled, such as impacts of changes in vegetation, albedo, or building
The Phoenix Urban Form Project outlined a new development plan to create a “connected oasis” that allows for street vibrancy, hospitality, pedestrian comfort, and interesting landscape. SOURCE: Studio Ma, http://studioma.com/.
type on meteorological endpoints. For example, results of small pilot programs—such as a pilot white roof heat island mitigation program in New York City—examining the impact of high-albedo white roofs to reduce surface temperatures, shows promise for large scale applications (Gaffin et al., 2012; Figure 2.4). Modeling the impacts of large-scale applications of such scenarios (Akbari et al., 2009; Oleson et al., 2010a) is important (as well as addressing the practical aspects of keeping the “white” surfaces clean so that they function as intended) to better understand how these applications may affect the atmosphere. Additional scenarios that could be modeled include higher-albedo land use changes at grade (e.g., pavements), although the technological challenges associated with land use changes at grade are greater than for roofs.
In addition, it appears end user groups concerned with sustainability are not yet confident in urban meteorologists’ ability to accurately model the complex effects that impact quality of life in the urban environment of cities.
FIGURE 2.4 The pilot white roof heat island mitigation program in New York City. Results of white vs. black roof surface temperatures. While both colors of roof fluctuate with day and nighttime temperatures, the white roofs stay consistently cooler during the daytime. (www.nyc.gov/coolroofs) SOURCE: Gaffin et al., 2012.
For example, complex models that consider street, road and building impacts have not consistently been able to improve on simple model forecasts.
Coupling of Urban Meteorological Models
with Other “Cascading Events”
Many workshop participants stated that development of coupled models with meteorological and other parameters would be particularly relevant for end users.
Coupled land-ocean-urban effects. Given that a majority of our large cities are in coastal regions, coupled land-ocean-urban effects are becoming increasingly important. The land-urban region does not act independently from the neighboring ocean, and vice versa. There are complex interactions and feedbacks within the coupled environment. Several recent papers highlighting the impact of air-sea interaction processes such as urban heat islands and sea breeze fronts on the urban environment illustrate this point (see Carter et al., 2011; Holt et al., 2009; Thompson et al., 2007). This work is complex in that it may involve the coupling of different modeling frameworks (e.g., NWP, urban canopy models, dispersion-transport, etc.). Coupling of computational fluid dynamics models (CFD) with Numerical Weather Prediction (NWP) models and fast-response emergency management (EM) tools is a good example of this (Boris et al., 2011).
Coupled atmospheric chemistry and meteorological models. A better understanding of gradients in ambient air pollution concentrations and their variation through coupled atmospheric chemistry and meteorological models could be useful to many end user groups. Improved particulate matter predictions could result in increased accuracy and spatiotemporal resolution and better representation of the diversity of (natural, anthropogenic, primary and secondary) sources and risks.
Applied weather forecasts for energy demand. As sustainable development efforts lead to more substantial eco-district3 (i.e., neighborhood scale) level energy management systems, such as district heating/cooling4 and electric vehicle connection to grid, there will be an increased need for management products that merge local scale weather forecasts with local scale energy demand and renewable supply availability forecasts. Such products
3 A concept that aims to reach sustainability goals through action at a neighborhood level.
4 A method of delivering heating and cooling to surrounding buildings in an eco-district using a highly efficient energy plant.
would allow for optimum use of resources in anticipation of weather events. For example, buildings could be precooled prior to a heat wave, warming shelters preheated before a cold wave, and electric vehicles could be used as energy storage to meet summer daytime peak load needs.
Applied weather and climate forecasts for public health. From a public health perspective, high temperatures influence the risk of cardiovascular and respiratory illness and infectious diseases (Patz et al., 2005). Accurate seasonal forecasts may enable public health systems to rapidly monitor, identify, and respond to new climate change and weather related health risks. Accurate forecasting of inundation areas for evacuation during coastal storms would also enable the public health and emergency response systems to respond more effectively.
Several areas were identified at the workshop where urban meteorological forecasting and monitoring capabilities are currently underutilized. For example, a recent American Meteorological Society report (AMS, 2011) found that drivers can make safer decisions about their travel plans and react appropriately during potentially compromised conditions when high quality weather information, including both current observations and forecasts, has been communicated to them in a timely and effective manner. However, there are several technical, financial, societal, and institutional barriers as discussed below that need to be overcome before the full potential of mobile observations can be realized by the weather and transportation communities.
Although forecasting and monitoring capabilities may exist, there appears to be a lack of accessible products, such as weather applications for smartphones; websites with information relevant to special urban problems such road hazards, flash floods, and drainage problems; and neighborhood-level alerting for areas with particular susceptibility to excessive heat and pollution hot spots. Several workshop participants commented that precision urban weather products are needed to provide personalized, local information that is readily accessible and easy to understand. The creation and provision of these types of products requires close, ongoing collaborations between urban meteorologists and departments of transportation, public health agencies, centers for disease control, urban planners, and others.
Probabilities may also be important to consider with respect to providing longer forecast lead times (see sections above on short-term weather and
long-term prediction). Probabilistic forecasting is a major advance in the field of urban meteorology, but to date it has been used primarily for research applications and not for other end users. There is an underutilization of “lead time” and uncertainty/probability information that operational centers can provide. This is because of the need for greater computing power, but also because the end users may not be educated on how to properly interpret model-generated probabilistic information. Similarly, the modelers may not know how to best communicate probabilistic information. As such, provision of probabilistic information will need to be accompanied by better communication of uncertainties to the general public (NRC, 2006).
Improved intra-and extra-community communications on resources, capabilities, and current knowledge is a significant crosscutting issue. The lack of such communications is also one of the reasons for underutilized urban forecasting and monitoring capabilities discussed in the previous section.
Current State of Communication and
Exchange Between Communities
In some cities, substantial amounts of data are being generated by urban meteorologists, and a host of tools are available for analysis. However, these resources are not being fully utilized by end users. Workshop participants identified several reasons for the lack of exchange between communities.
Lack of cross-fertilization and cross-discipline training. The models and capabilities may exist at the urban meteorological end, but local management and operations staff are not trained to use the capabilities; they may not even know what they can obtain from the various capabilities. A common theme of “we don’t know what we don’t know” (regarding existing capabilities) was present at the workshop. Successful cross-fertilization typically involves the urban meteorologist and the end user communicating with each other on a regular basis, leading to a greater understanding of capabilities on both sides.
User needs are not being translated to the urban meteorological community. Exchanges between urban meteorologists and end users need to be encouraged during project planning phases, so that outputs from the urban meteorological community are tailored to user needs (see Section 1). There are important limitations with respect to resources and funding to make
such collaborations or discussions successful and worthwhile. For instance, currently, many U.S. cities do not have validated health-based criteria for issuing heat advisories, watches and warnings. The “Heat Health Watch Warning System,” based on synoptic weather classification5 promoted by the NWS, has not been demonstrated to be superior to setting appropriate health-based criteria for simpler weather metrics (e.g., Metzger et al., 2010).
Constraints on end users (i.e., institutional barriers). Due to the nature of operations and funding sources, procurements and services requested by certain groups and sectors are governed by various laws, regulations, agreements, and manuals, thus possibly limiting interactions with relevant representatives of the urban meteorological community. It is important for this community to acknowledge that meteorological information is just one of many inputs into most end users’ considerations and decision-making.
These apparent barriers can be overcome by establishing better relationships at the local level. To gain a better understanding of the different constraints by users, end-to-end dialogue as well as end-to-end-to-end dialogue could be developed between end users and urban meteorologists. An end-to-end-to-end dialogue consists of multidirectional communication and sustained interactions among researchers, application developers, and multiple decision makers and is done over several iterations to coproduce information that is useful for various societal applications (Morss et al., 2005). There are numerous opportunities for the urban meteorological community to interact with the end user community both locally and regionally, especially during the mitigation and preparedness phase of planning. The end user community also has a responsibility to reach out and understand the resources and capabilities that exist within the urban meteorological community.
Environmental (and other) data are ultimately a commodity, as discussed in the section on Data Access and Data Sharing Needs.
The Need for Strengthened and Improved Communication
The ultimate goal of the applied science of meteorology is to help society—both in the collective sense and as individuals—better anticipate and take appropriate response to an ever-changing environment or an imminent risk. The effectiveness of improved science is measured as much by society’s response to the new information as the quality of the science itself. Weather impact forecasts are needed that communicate hazards and their impacts on people and infrastructure, not simply the meteorological
conditions. To create and communicate weather forecast information that is usable to end users in ways that inform their decision-making requires social science research (conducted in partnership with meteorologists) and ongoing relationships with end users.
Several workshop participants noted that it is crucial that the science of risk perception and risk communication be further applied to evaluate and improve effective dissemination of urban weather and climate risks to diverse urban populations and policy makers. Failure to do so could undermine all the upstream science, observations, and modeling.
A great deal of discussion at the workshop revolved around communicating information specific to urban meteorology and its impacts on the general public, including real-time hazard releases, observations of flooding, and neighborhood-level excessive heat and pollution advisories, as well as short-term forecasts for hazards). While this report cannot discuss the finer points of communication practice in detail,6 most presenters at the workshop made note of important communication-related needs that can be defined broadly as fitting within risk communication. Typically, risk communication refers to the set of communications conducted prior to the incidence of a threat (often designed to avoid a crisis).
Threat perception differs by the type of threat and by the perspective of those assessing the threat to themselves and their operations. For example, most deaths and illnesses caused by excessive heat cannot be directly observed or attributed on a case by case basis. The risks posed by heat waves, therefore, may be underestimated by both mass media and the general public compared to other natural or manmade disasters (e.g., fires, floods, plane crashes). It is important that the general public identifies the risks as well as how to respond to them. Sharing anecdotes and stories from affected people might be a more effective way of communicating risk than statistical estimates of public health impact. There is also a need for both urban meteorologists and emergency managers to work with communication experts to identify better ways of communicating risks.
Good communication practices necessitate establishing relationships prior to the incidence of a hazard, establishing credibility, and providing for expedient information transfer, as well as utilizing longer lead-times with greater spatial resolution prior to the incidence of hazards (i.e., providing for mitigation). If urban weather products from multiple vendors, whether private or public, were recognized by end users to be in conflict, the
6 See NRC (2006) for more information on ways to improve the generation, communication, and potential use of uncertainty information for hydrometeorological forecasts.
credibility of the corpus of information presented could be diminished and actionability could be reduced.
The end goal of successful communication of usable forecast information is the protection of life and property, as well as enhancing economic outcomes for private and public sectors. Including these metrics is important for gauging the success of urban meteorology. It is important to highlight socioeconomic benefits, but it should also be stressed that minimizing negative impacts is as important as maximizing direct positive outcomes. This demands that physical, health, and social scientists need to work far more collaboratively than in the past.
It is important to recognize that most of the discussions here are related to communication practices. Communication is also a social scientific field of study with many subfields. Subfields relevant to urban meteorology and end users include risk communication (as noted), science communication, organizational communication, and mass communication. There are myriad theories, concepts, and lessons from past research that can be applied to the urban meteorology context, and future research would further develop these concepts and theories.
Advanced Approaches to Training and Workforce Development
Numerous workshop participants asserted that advanced approaches to training and workforce development will be critical in producing the next generation of urban meteorological models and applications. In particular, interdisciplinary approaches are increasingly needed to address critical gaps in the interface between natural, biological, and human systems in the urban landscape. In addition to promoting interdisciplinary team approaches, it is becoming increasingly necessary to cross-train scientists in multiple disciplines. Widespread training of meteorologists about the unique weather and climate features of the urban environment is also important (Grimmond abstract, Appendix A).
To strengthen communication, it should be fostered across scientific disciplines and spatial and temporal timescales (Grimmond abstract, Appendix A). It is crucial that urban meteorologists and end users attend joint conferences (such as the workshop for this report) and each other’s professional conferences. It is also mutually beneficial to jointly train graduate students and postdoctoral researchers. For instance, a new postdoctoral training program between the National Center for Atmospheric Research
(NCAR) and the Centers for Disease Control and Prevention (CDC) requires the candidate to stay at NCAR for one year to become familiar with weather and climate prediction, followed by another year at CDC to integrate urban meteorology with public health.
It is also mutually beneficial to jointly train graduate students and postdoctoral researchers as well as exchanging personnel. This will give all groups involved a better understanding of what is available and unavailable in urban meteorology and how and what urban meteorology information is used in decision-making by end users. Promoting the use of interns to work with organizations may be a feasible option; such a process may help to cross-pollinate communities. It may also be beneficial to create funds for end users to undergo training with research groups; existing programs tend to only educate young and new students. Grant programs for senior operations managers would be valuable in supporting their interactions with researchers in the field so that they would not need to rely on specialized contractors in order to stay current. There are also approaches in place by several institutions that require and allow for interdisciplinary discussion, such as the Environmental Protection Agency’s (EPA) Requests for Applications7 targeting community-based participatory research and the National Oceanic and Atmospheric Administration’s (NOAA) Regional Integrated Sciences and Assessments (RISA) program.8
Design Strategies for Communicating Data
and Results Between Communities
Geographic Information Systems (GIS) have become a common software tool across disciplines and communities. GIS-based tools allow for relatively easy data visualization and provide a common data management approach that aids in data sharing among disciplines. Repeatedly, stakeholder communities have requested data in formats like GIS and Keyhole Markup Language (KML) rather than discipline-specialized, complex formats (e.g., hdf, netcdf). Emerging technologies such as agent-based modeling, geo-visualization, mobile GIS, and spatial statistics are also powerful resources for expanding the reach and utility of urban meteorological data, decisions, and warnings.
The emergence of real-time social media (e.g., Twitter, Facebook, YouTube) provide new formats for dissemination, warnings, and research within
the urban environment (NRC, 2011a). The notion of “social sensors,” in which social networking users are considered as “sensors” and the messages that they share are considered “sensory information,” has recently been discussed in regard to other natural hazards (Sakaki et al., 2010) and would certainly offer opportunities in urban meteorological forecasting and hazard alerts. Other common dissemination strategies include traditional media (e.g. television, radio, and internet), text message alerts, and telecommunications. However, along with timely communication facilitated by social networks comes a risk of false information spreading rapidly. There is thus a need to use social media actively for dissemination of information and at the same time also to screen/alert to possibly false information posted via unreliable sources on these networks.
Design Strategies for Data Sharing Between Communities
As mentioned earlier, many federal, state, and city government agencies typically collect environmental data that complement their operational needs. For various reasons, whether legal or institutional, that data is often not shared externally. However, such data may be particularly useful for research and operations in urban meteorology, which could potentially enhance the value of the data for the agency from which it originated. For example, if an urban drainage and wastewater utility shared stream flow and built environment information, it may allow operational meteorologists to disseminate better warnings, and it may enhance the modeling capabilities of research meteorologists. Advancements in such data sharing are likely to have significant impacts to public safety and welfare in urban areas. Conversely, urban meteorology researchers could provide end users with metadata such as raw output, model parameterization and/or biases; highlighting areas for model improvement may invite greater sharing from end users and agencies. An example of effective two-way flow of information is the collaboration between Seattle Public Works and the National Weather Service (NWS), where the utility obtains NWS data and returns water flow data to NWS (see Box 2.4).
Although there are considerable logistical and budgetary constraints, a fully resourced data center could facilitate data sharing. A database system could be created to receive data in real time to allow [approved] users to access it in real time. It could facilitate a number of agencies in meeting their objectives (e.g., rainfall data for flood advisories) and allow wider use of current observations and model products. Ideally the system would have capabilities that include the following:
1. Accepting real-time and older data that undergo quality assurance/ quality control
2. Receiving multiple formats of data
3. Accepting metadata (siting, instrument characteristics, sampling) that are regularly updated
4. Developing algorithms that can provide additional information in real time to data suppliers (e.g. outliers, data source not online) that provide extra value for those who are supplying the data
5. Creating a database that can be queried in real time and used for re-analysis
6. Providing modeling tools or model outputs that use the observational data and provide modeling data products that can be used by other end users (e.g. meteorological model outputs that could be used by transport modelers, air quality modelers)
The intention would be to build a smart data system that would enhance the use of current measurements. The model output database would allow development of model ensembles.
Interdisciplinary Urban Testbeds
Transitioning Research to Operations
There is an increase in numbers of field programs and testbeds developed specifically to explore urban issues. Some modelers are considering using the data from the urban testbeds to examine quality of life issues. These testbeds provide long-term observations and models for city-specific input into planning (Figure 2.5). They have not necessarily been stakeholder driven from the start, but they have presented opportunities for stakeholders to participate and contribute to project design and outcomes (see Chapter 4 for detailed discussion).
A significant focus of the workshop was to discuss the needs of users of urban meteorology information and brainstorm possible solutions and strategies to better meet user needs. Several key themes emerged from this discussion.
FIGURE 2.5 Locations and pictures of various instruments of the NYCMetNet. (A) Temperature, humidity and liquid water vertical profiler (to 2 km). (B) and (F) Sodar wind profiler to 300/450 m. (C) Radar wind profiler vertical profiler (to 2 km). (D) CCNY Aerosol Raman lidar (to 10 km) and Vaisala ceilometer. (E) Skyscraper-mounted weather stations. Colored arrows are located at the position where weather stations are located and the information is continuously ingested and archived via the NOAA MADIS system. The arrows represent measurement of wind speed and direction. Not shown is a portable eye safe Doppler Lidar, a network of radiation flux observing instruments, a Nephalometer and other particulate matter measurement stations to sample ground level aerosols for comparison to the vertical profiling instruments. SOURCE Mark Arend, CCNY Optical Remote Sensing Lab and NOAA CREST.
• A clear mechanism to help the urban meteorological community better identify user groups, reach out to them, and begin an ongoing dialogue to assess and better meet their needs has yet to be identified.
• It was apparent at the workshop that end users of urban meteorological information are heterogeneous and cover a vast spectrum of job roles, goals, needs, experience, and understanding. Acknowledging and understanding this heterogeneity is important for the urban meteorological community to better understand, interact with, and meet the needs of end users.
• Strong communication early in the process and throughout is key to successful coordination. This is a common theme present in several case studies, and examples discussed in this chapter highlighted success stories where the needs of the end user were met to a large degree (see Boxes 2.3, 2.6). In some cases, the end users worked directly with researchers to develop the tool they required (see Boxes 2.3, 2.4), which helped ensure they received a tool that could be tailored to their specific needs. The hallmarks of effective communication are a better understanding of capabilities on both sides; the successful translation (through collaboration) of user needs to the research community and research products to the end user community; and a better understanding in the research community of institutional constraints that end users face.
• Many workshop participants noted that more coordinated data sharing strategies among various agencies and training of various end user communities could help them utilize existing data. Several other data sharing issues are lacking, including quality assurance and metadata needs that could be addressed by the research and weather services communities as well as the end users. Coordinated data formats, available metadata, and novel dissemination strategies are increasingly essential as urban meteorological data and model output cross the research-to-operations “valley.”
• The workshop revealed an underrepresentation of urban meteorological, climatological, and field coursework and training at all educational levels. Advanced approaches to training and workforce development will be critical in producing the next generation of urban meteorological models and applications. It is important that students and professional stakeholders continue to be properly educated as the complexities of the urban meteorological/climatological environment evolve and as the academic community approaches cities as coupled human and natural systems that need to be sustainable. To strengthen training, it is crucial that urban meteorologists and end users attend joint conferences (such as the workshop for this report) and each other’s professional conferences. It is also mutually beneficial to jointly
train graduate students and postdoctoral researchers as well as exchanging personnel. This will give all groups involved a better understanding of what is available and unavailable in urban meteorology and how and what urban meteorology information is used in decision-making by end users.
Above all, it is essential that the urban meteorological community understands what data are needed by end users that are not currently produced or not conveyed in usable ways to end users. Creating and communicating weather forecast information that is usable to end users in ways that inform their decision-making requires social science research (conducted in partnership with meteorologists) and ongoing relationships with end users.