Albuquerque

New Mexico

s t ra t e g i e s Report
 s t ra t e g i e s

The CAPA Heat Watch program, equipment, and all related procedures referenced herein are
developed through a decade of research and testing with support from national agencies and
several universities. Most importantly, these include our partners at the National Integrated Heat
Health Information System, the National Oceanic and Atmospheric Administration’s (NOAA’s)
Climate Program Office, and National Weather Service, including local weather forecast offices at
each of the campaign sites, The Science Museum of Virginia, and U.S. Forest Service (USDA). Past
support has come from Portland State University, the Climate Resilience Fund, and the National
Science Foundation. We are deeply grateful to these organizations for their continuing support.

This report was prepared by CAPA Strategies, LLC

Summer 2021
 Credit: Jesse Wood

Table of Contents
4 Executive Summary

5 Purpose and Aims About the Maps

6 Campaign Process 8 Initial Observations

9
7
Morning Traverse Points
Maps
10 Morning Area-Wide
13 Mapping Method
11 Afternoon Traverse Points

14 Heat Modeling Summary 12 Afternoon Area-Wide

15 Next Steps

16 Media
 Executive Study Date
s t ra t e g i e s
Summary
July 9th, 2021
Major thanks to all of the participants and organizers of the Urban
Heat Watch program in Albuquerque, New Mexico. After months
of collaboration and coordination, local organizers and volunteers 65
collected thousands of temperature and humidity data points in Volunteers
the morning, afternoon, and evening of a long, hot campaign day
on July 9th, 2021.
18
Routes
Morning Area-Wide Temperature (6 - 7 am) Afternoon Area-Wide Temperature (3 - 4 pm)
622.2°F 79.1°F 944.4°F 105.3°F
67,662
Measurements

105.9°
Max
Temperature
N N

16.9°
120 120 Temperature
min mean max

min mean max

Differential
100 100

80 80

60 60
Learn more about the background and goals
6-7 3-4 of each Heat Watch 2021 campaign city at
am Traverse Point Temperature pm https://nihhis.cpo.noaa.gov/Urban-Heat-
Islands/Mapping-Campaigns/Campaign-Cities
 Purpose & Aims
We know that climate-induced weather events have the most
profound impact on those who have the least access to financial
resources, historically underserved communities, and those
struggling with additional health conditions. Infrastructure is also
at risk, which can further compromise a region’s capacity to
provide essential cooling resources.

CAPA Strategies offers an unparalleled approach to center

communities and infrastructure facing the greatest threat from
the impact of increasing intensity, duration, and frequency of
extreme heat. This report summarizes the results of a field

1
campaign that occurred on July 9th, 2021 and
with it we have three aims:

2
Provide high resolution
descriptions of the
distribution of tempera-
ture and humidity (heat
index) across an
urban area Engage local communities
and create lasting partner-
ships to better understand
and address the inequitable

3
threat of extreme heat

Bridge innovations in sensor

technology, spatial analytics,
and community climate
action to better understand
the relationships between
With a coordinated data-collection campaign over
urban microclimates, infra- several periods on a hot summer day, the resulting data
structure, ecosystems, and provide snapshots in time of how urban heat varies
human well-being. across neighborhoods and how local landscape features
affect temperature and humidity.
5
 1. Set Goals
Campaign organizers determine the extent of their
mapping effort, prioritizing areas experiencing
environmental and social justice inequities. CAPA
then divides this study area into sub-areas
(“polygons”), each containing a diverse set of land
uses and land covers.

Campaign Process 2. Establish

Organizers recruit volunteers, often via non-profits,
CAPA Strategies has developed the Heat Watch universities, municipal staff, youth groups, friends,
campaign process over several iterations, with family, and peers. Meanwhile, CAPA designs the
data collection routes by incorporating important
methods well established through peer-re- points of interest such as schools, parks, and
viewed publications¹, testing, and refinement. community centers.

The current campaign model requires leader- 3. Prepare

ship by local organizers, who engage communi- Volunteers attend an online training session to learn
ty groups, new and existing partner organiza- the why and how of the project, their roles as data
tions, and the media in generating a dialog collectors, and to share their personal interest in the
project. Participants sign a liability and safety
about effective solutions for understanding and waiver, and organizers assign teams to each
addressing extreme heat. polygon and route.

CAPA provides training, equipment, and support

to the recruited community groups as they 4. Activate
endeavor to collect primary temperature and With the help of local forecasters, organizers
humidity data across a metropolitan region. identify a high-heat, clear day (or as near to one as
possible) and coordinate with their volunteer teams.
Once confirmed, CAPA ships the sensor equipment
The seven main steps of the campaign process and bumper magnets to be distributed to campaign
participants.
are summarized to the right. An overview of the
analytical modeling methodology is presented
later in this report and described at full length 5. Execute
in peer-reviewed publications. Volunteer teams conduct the heat campaign by
driving and/or bicycling sensor equipment along
pre-planned traverse routes at coordinated hour
intervals. Each second the sensors collect a
measurement of ambient temperature, humidity,
longitude, latitude, speed and course.

¹ The most relevant and recent publications to the 6. Analyze

Heat Watch campaign process include: Organizers collect and return the equipment, and
CAPA analysts begin cleaning the data, as described
Shandas, V., Voelkel, J., Williams, J., & Hoffman, in the Mapping Method section below, and utilize
machine learning algorithms to create predictive
J., (2019). Integrating Satellite and Ground
area-wide models of temperature and heat index for
Measurements for Predicting Locations of each traverse.
Extreme Urban Heat. Climate, 7(1), 5. https://-
doi.org/10.3390/cli7010005

Voelkel, J., & Shandas, V. (2017). Towards 7. Implement

Systematic Prediction of Urban Heat Islands: Campaign organizers and participants review the
Grounding Measurements, Assessing Modeling Heat Watch outputs (datasets, maps, and report),
Techniques. Climate, 5(2), 41. https://- and campaign teams meet with CAPA to discuss
doi.org/10.3390/cli5020041 the results and next steps for addressing the
distribution of extreme heat in their community.
6
 About The Maps
The following sections present map images from the Heat Watch
campaign and modeling process. Two sets of maps comprise the final
results from the campaign process, and they include:

Point temperatures collected in each Area-wide heat maps, displaying either the
traverse period, filtered to usable data. modeled temperature or heat index across the
entire study area at each traverse period.

80.2 F 97.2 F

Warmer Cooler

73.1 F 83.3 F

The data are classified by natural breaks in order Note that the scales are different between
to clearly illustrate the variation between warmer the traverse point and area-wide maps due
(red) and cooler (blue) areas across the map. to the predictive modeling process.

How does your own experience with heat in these areas align with the map?

Find your home, place of What about the landscape

work, or favorite park on the (trees, concrete buildings,
maps and compare the heat riverside walkway) do you
throughout the day to your think might be influencing
personal experience. the heat in this area?

7
 Initial Observations
The distribution of heat across a region often varies by qualities of the
land and its use. Here are several observations of how this
phenomenon may be occurring in your region.

Trails abbutting vegetation and

arroyos offer cooler paths for
pedestrians and bicyclists than
the busy intersecting highways
and car-filled streets.

Streets nearby to green space and dense

tree canopy show cooler temperatures.
Wide asphalt intersections with little to
no shade retain high temperatures and
offer no refuge for pedestrians.

8
 Morning Traverse Points
(6 - 7 am)

≤66.9°F ≤68.3°F ≤69.6°F ≤70.6°F ≤71.5°F ≤72.4°F ≤73.3°F ≤74.3°F ≤76.2°F

Model Boundary

0 2 4

miles
9
 Morning Area-Wide Predictions
Temperature (6 - 7 am)

62.2°F Model Boundary 78.4°F

0 2 4

miles
10
 Afternoon Traverse Points
(3 - 4 pm)

≤95.8°F ≤96.8°F ≤97.4°F ≤98.1°F ≤98.7°F ≤99.4°F ≤100.1°F ≤101.0°F ≤103.0°F

Model Boundary

0 2 4

miles
11
 Afternoon Area-Wide Predictions
Temperature (3 - 4 pm)

94.4°F Model Boundary 105.3°F

0 2 4

miles
12
 Mapping Method

1
Download &
Filter
Download raw heat data Compare data with field Trim data to proper time
from sensor SD cards notes and debrief interview window, speed, and study area

2
Integrate &
Analyze
Download multi-band Transform land cover Calculate statistics of
land cover rasters from rasters using a moving each land cover band
Sentinel-2 satellite window analysis across multiple radii

3
Predict &
Validate
Combine heat and land Create predictive Perform cross validation
cover data in Machine raster surface models using 70:30 holdout
Learning model of each period method

The most relevant and recent publications include:

Shandas, V., Voelkel, J., Williams, J., & Hoffman, J., (2019). Integrating Satellite and Ground Measurements for
Predicting Locations of Extreme Urban Heat. Climate, 7(1), 5. https://doi.org/10.3390/cli7010005
Voelkel, J., & Shandas, V. (2017). Towards Systematic Prediction of Urban Heat Islands: Grounding
Measurements, Assessing Modeling Techniques. Climate, 5(2), 41. https://doi.org/10.3390/cli5020041

13
 Credit: Gino Barasa

Accuracy Assessment*

Traverse R-Squared
6 - 7 am 0.98
3 - 4 pm 0.95

Field Data
Like all field campaigns, the collection of temperature and humidity data requires carefully following provided
instructions. In the event that user error is introduced during the data collection process, outputs may be compromised
in quality. While our team has a developed a multi-stage process for assessing and reviewing the datasets, some
errors cannot be identified or detected, and therefore can inadvertently compromise the results. Some examples of
such outputs may include temperature predictions that do not match expectations for an associated landcover (e.g. a
forested area showing relatively warmer temperatures). We suggest interpreting the results in that context.

Prediction Areas
The traverse points used to generate the areas wide maps do not cover every square of the studied area. Due to the
large number of data collected, however, our predictive models support the extension of prediction to places beyond the
traversed areas. We suggest caution when interpreting area wide values that extend far beyond the traversed areas

*Accuracy Assessment: To assess the strength of our predictive temperature models, we used a 70:30 "holdout cross-validation method," which
consists of predicting 30% of the data with the remaining 70%, selected randomly. An 'Adjusted R-Squared’ value of 1.0 is perfect predictability,
and 0 is total lack of prediction. Additional information on this technique can be found at the following reference: Voelkel, J., and V Shandas, 2017.
Towards Systematic Prediction of Urban Heat Islands: Grounding measurements, assessing modeling techniques. Climate 5(2): 41.
14
 Next Steps

ow that you have completed a Heat Watch Growing Capacity services reflect a holistic
campaign, you have a better understand- approach to climate change mitigation and
ing of where urban heat is occurring in adaption. Our process is rooted in social scientific
your region, and who is at risk of exposure. You thinking, interdisciplinarity, and a mission of
may be wondering what to do next: how to equity. This adds up to capacity-building solu-
mitigate that exposure, or help your region adapt tions which are actionable, tailored to your
to a hotter future. If you would like to take the region, and promote climate resilience for all.
next steps in preparing for climate change,
CAPA’s Growing Capacity services can help. We offer a range of services to support you in
your climate adaption efforts, no matter how big
or small. Choose from our offerings below to
create a Growing Capacity package that fits your
needs and budget.

Growing Capacity is an arm of CAPA Strategies

which emphasizes place-based solutions,
substantive community engagement, and the
translation of data into action. These services ask
not only “where do climate risks exist?,” but
“what can we do about them?” Growing Capacity
services offer a systematic way to integrate data
and accelerate climate adaptation in your area.
We do this by reducing common barriers that Whether your climate adaption goals require
limit action; making climate adaptation accessi- increased community-based research, data
ble to your colleagues and communities; and synthesis, public outreach, network-building, or
facilitating opportunities for collaboration, learn- novel interventions, the Growing Capacity team
ing, and problem solving. is here to assist you.

Jurisdictional Capacity Community Resource

Scan Assessment Knowledge Development
Assessment
Comprehensive Comprehensive Strategic plans,
Report Report, Analysis Handbooks, Policy
Workshops,
Surveys, Focus language, Tools for
Groups, Interviews education/outreach
Want to start a conversation about Growing Capacity
in your region? Contact us at info@capastrategies.com 15
 Media

@capa_heatwatch @capaheatwatch

www.capastrategies.com 16
s t ra t e g i e s