Skip to content U.S. Department of Transportation/Federal Highway AdministrationU.S. Department of Transportation/Federal Highway Administration

Office of Planning, Environment, & Realty (HEP)
PlanningEnvironmentReal Estate

HEP Events Guidance Publications Awards Contacts

Climate Variability and Change in Mobile, Alabama

3. Temperature

Temperature can affect transportation infrastructure in many ways and is a key consideration in many transportation designs, maintenance schedules, and operation budgets. Most importantly, extreme temperature events can damage transportation infrastructure.1 Depending on the severity and duration of an extreme temperature event, bridges, pavement, vehicles, and construction activity are at risk.2 For example, thermal expansion of paved surfaces can cause infrastructure degradation.3

This section presents the methodology and key findings for several analyses related to observed and projected temperature.

Additional detail about the temperature analyses is available in the appendices.

3.1. Observed Temperature

Observed temperature records for Mobile were analyzed to describe historical and present-day climate conditions in the region.

3.1.1. Methodology

This section describes the methodology used to analyze observed temperature data in the Mobile region.

Historical data from five National Oceanic and Atmospheric Administration (NOAA) Global Historical Climatology Network (GHCN) stations in the Mobile region were analyzed to investigate existing climatic trends and baseline conditions.6 Two of the stations (Coden and Mobile Airport) are located in Mobile County and three of the stations are located in neighboring Baldwin County (see Figure 4).7 Table 6 summarizes the data available for each station and identifies any gaps in the record. Temperature observations at the Baldwin County stations began between 1915 and 1924, giving these stations the longest record of temperature measurements. Temperature observations at the Mobile County stations began between 1948 and 1956.

Figure 4: Map of GHCN Stations

This figure shows a map of the five NOAA Global Historical Climatology Network (GHCN) stations in the Mobile area that were analyzed to investigate existing climatic trends and baseline conditions. Two stations are located in Mobile County. Mobile Airport is directly west of downtown Mobile and Coden is in southern Mobile County, close to the Gulf Coast. The remaining three stations are in Baldwin County. Bay-Minette is in the northeast area of the study region, Fairhope is just to the east of Mobile Bay, and Robertsdale is in eastern Baldwin County.
Table 6: Mobile Area GHCN Station Temperature Data Summary
Station Station ID# County Temp. Start of Data Collection Temp. Data Gaps*
Bay-Minette USC00010583 Baldwin 1915 1931, 1937-1941
Coden USC00011803 Mobile 1956 1985-1988
Fairhope USC00012813 Baldwin 1918 -
Mobile Airport USC00015478 Mobile 1948 -
Robertsdale USC00016988 Baldwin 1924 1926-1934

* Gaps defined as over 80% of data points missing for the year
** Missing minimum temperature only

Each station records daily minimum and daily maximum temperatures. Daily mean temperature was estimated by averaging the daily minimum and maximum temperatures.8 Monthly, seasonal, and annual minimum, maximum, and mean temperature averages were produced using the daily data for both the historical record (1912 to 2009), where data were available, and the present-day climate period (1980 to 2009). This present-day climate period (1980 to 2009) corresponds to the baseline climate period adopted for the climate projections. A Mann Kendall trend analysis was used to identify statistically significant changes in annual and monthly minimum, maximum, and mean temperatures from 1961 to 2009.2 Additional analyses were conducted to illustrate historical and present-day conditions of temperature extremes.

3.1.2 Key Findings

This section describes key findings from the analysis of observed temperatures. Key findings are presented for annual, seasonal, and monthly temperatures, and temperature extremes.

Key Findings for Temperature Trends

  • Average annual maximum temperature in the Mobile region from 1912-2009 was 78°F (25°C), with the hottest monthly temperature experienced in July at 90°F (32°C).
  • Overall, average annual temperatures have not changed significantly in the Mobile region.
  • Average maximum and mean temperatures in March and September have decreased significantly in the Mobile region.
  • Average annual maximum temperatures have decreased significantly at three of the GHCN stations.
  • The number of days per year below freezing ranged from 6 days in 1921 to 43 days in 1978 and averaged 23 days over the entire historical record.
Annual Temperatures

The Mobile region stations have experienced similar patterns in temperature over time, collectively demonstrating no regional statistically significant (p<0.10) trends in annual temperatures over the twentieth century.3 Averaging across all stations and all years, 1912 to 2009, the average annual temperatures were as follows4:

The Coden station tends to be the coolest of the five stations, with an average observed mean temperature of 66°F (18.9°C), or approximately 0.8°F (0.4°C) lower than the five-station average.

Table 7, Table 8, and Table 9 provide the average annual minimum, maximum, and mean temperature, respectively, for each station over both the full station record (see Table 6 for each station's full record), the most recent climate period (1980-2009)5, and the historical record used to inform the trend analysis (1961-2010).6 The tables also provide the standard deviation (SD) of the temperature values across the time series. The SD values indicate how much variability exists in the data. A large SD as a proportion of the mean indicates a large amount of variability in the historical values, while a smaller SD indicates that the values do not vary considerably over the time period. The Coden and Robertsdale stations seem to experience the greatest temperature variability. It is not known if this variability is, in fact, due observational error.

The fourth column for each station in Table 7, Table 8, and Table 9 notes whether there has been a statistically significant increase or decrease in temperature from 1961 to 2010.7 Temperatures have not changed consistently in the Mobile region. The most consistent trend found is that average annual maximum temperatures have decreased significantly (p<0.01) at three of the GHCN stations, though no trend was observed at the Fairhope and Mobile stations. These findings are fairly consistent with observations in the Southeast (see Appendix B.1 for more information).

The only temperature variables that have shown regionally consistent statistically significant trends over the data record are the average maximum and mean temperatures for March and September, which have decreased significantly (p<0.01) at all five GHCN stations.

Table 7: Average Annual Minimum Temperature (°F) and Variability (?°F) for Each Station
Station Full Station Record+ 1980-2009 1961-2010 Historical Trend?
Bay-Minette 56.5 (1.7) 56.5 (0.8) 56.4 (1.1) decreasing***
Coden 55.5 (4.2) 54.7 (5.2) 55.6 (4.3) no trend
Fairhope 57.3 (1.4) 56.9 (0.9) 56.9 (0.9) no trend
Mobile 57.4 (1.0) 57.1 (0.9) 57.4 (1.0) no trend
Robertsdale 54.0 (5.6) 55.5 (1.4) 55.3 (1.4) decreasing**
Average 56.1 (2.8) 56.1 (1.8) 56.3 (1.7)  

+Full Station Record is detailed in Table 6
*Statistically significant at the: *90% Confidence Level, **95% Confidence Level, ***99% Confidence Level
Note: The standard deviation representing variability across the time period is provided in the parenthesis. Mann Kendall results exploring historical trend from 1961 to 2010 are provided courtesy of Dr. Hayhoe and Dr. Stoner of Texas Tech.

Table 8: Average Annual Maximum Temperature (°F) and Variability (?°F) for Each Station
Station Full Station Record+ 1980-2009 1961-2010 Historical Trend?
Bay-Minette 77.8 (1.8) 77.3 (1.1) 77.4 (1.0) decreasing***
Coden 76.5 (3.2) 75.1 (3.5) 76.4 (3.2) no trend
Fairhope 77.8 (1.7) 77.4 (1.2) 77.4 (1.2) decreasing**
Mobile 77.5 (1.0) 77.5 (1.0) 77.6 (1.0) no trend
Robertsdale 77.9 (1.9) 77.5 (1.4) 77.6 (1.3) decreasing*
Average 77.5 (1.9) 77.0 (1.6) 77.3 (1.5)  

+Full Station Record is detailed in Table 6
*Statistically significant at the: *90% Confidence Level, **95% Confidence Level, ***99% Confidence Level
Note: The standard deviation representing variability across the time period is provided in the parenthesis. Mann Kendall results exploring historical trend from 1961 to 2010 are provided courtesy of Dr. Hayhoe and Dr. Stoner.

Table 9: Average Annual Mean Temperature (°F) and Variability (?°F) for Each Station
Station Full Station Record+ 1980-2009 1961-2010 Historical Trend?
Bay-Minette 67.1 (1.7) 66.9 (0.8) 66.9 (0.9) decreasing***
Coden 66.0 (3.7) 64.9 (4.2) 66.0 (3.8) no trend
Fairhope 67.4 (1.4) 67.1 (0.9) 67.1 (0.9) decreasing**
Mobile 67.4 (0.9) 67.3 (0.8) 67.5 (0.9) no trend
Robertsdale 66.7 (1.6) 66.5 (1.3) 66.4 (1.2) decreasing**
Average 66.9 (1.9) 66.5 (1.6) 66.8 (1.5)  

+Full Station Record is detailed in Table 6
*Statistically significant at the: *90% Confidence Level, **95% Confidence Level, ***99% Confidence Level
Note: The standard deviation representing variability across the time periods is provided in the parenthesis. Mann Kendall results exploring historical trend from 1961 to 2010 are provided courtesy of Dr. Hayhoe and Dr. Stoner.

Seasonal and Monthly Temperatures

Figure 5 illustrates the average monthly and seasonal minimum and maximum temperatures across the historical record for each GHCN station. Average monthly mean temperatures range between 51°F (11°C) in January, the coldest month, and 81.2°F (27.3°C) in July, the warmest. This figure demonstrates there is more variability by month and season than by station location.

Figure 5: Average Monthly and Seasonal Minimum and Maximum Temperatures (°F), 1921-2009

Figure 5 illustrates the average monthly and seasonal minimum and maximum temperatures across the historical record for each GHCN station. Average monthly mean temperatures range between 51°F (11°C) in January, the coldest month, and 81.2°F (27.3°C) in July, the warmest. This figure demonstrates there is more variability by month and season than by station location.
Temperature Extremes

Temperature extremes have similarly demonstrated no significant changes over the historical data record. The highest recorded temperature was 106°F (41.1°C), recorded in 1925 at the Bay-Minette station. Averaged across all five stations, the hottest day of the year has ranged between 93°F (34°C) and 104°F (40°C), recorded in 2003 and 1930, respectively. The hottest day of the year has fluctuated around an average of nearly 98ºF (37°C), exhibiting no significant linear trend upward or downward over the data record (see Figure 6).15

Figure 6: Hottest Day of the Year (°F)

This figure plots a line graph of the temperature of the hottest day of the year at each of the five stations from 1912-2009, along with a 10-year moving average of the five stations. The temperatures fluctuate from year to year, but the moving average slows a slight decline over time. The moving average shows that the hottest day of the year has ranged between 96.5 and 99.9 degrees.

The number of days per year above 95°F (35°C) (see Figure 7) has shown a significant (p<0.1) decrease over the data record at three of the five GHCN stations from 1961 to 2009.

Figure 7: Number of Days per Year above 95°F

This figure shows a bar graph with the average number of days per year above 95 degrees at the five stations in the Mobile region, and the 10-year moving average of those values. The highest bar in the figure occurred in 1954, when there were 52 days above 95 degrees. The long-term moving average ranges between about 6 and 20 days per year.

As illustrated in Figure 8, the average coldest day of the year across the entire record and station locations was 19°F (-7°C). The coldest daily minimum temperature on record was 2.8°F (-16.2°C) in 1985. Over the past 30 years, the coldest day of the year for the Mobile region has become warmer, after reaching a historic low in the early 1980s.

Figure 8: Coldest Day of the Year

This figure shows the temperature of the coldest day of the year at each of the give stations from 1912-2009, along with a 10-year moving average of the five stations. The stations show similar values across the time series, but fluctuating from year to year. The moving average shows that coldest day temperatures were approximately 20 degrees until the 1950's, dipped in the 1960's and 1980's, and were just above 20 degrees in the late 90's and 2000's.

The number of days per year below freezing, averaged across all stations, ranged from 6 days in 1921 to 43 days in 1978 (see Figure 9) and averaged 23 days over the entire historical record.

Since 1960, the region has experienced an increase in the number of days below freezing compared to the first half of the century.

Figure 9: Number of Days per Year below Freezing

This figure shows a bar graph with the average number of days per year below freezing at the five stations in the Mobile region, and the 10-year moving average of those values. The moving average ranges between about 14 and 32 days per year above freezing. The moving average fluctuates over time, but appears to have an overall positive slope.

Table 10 provides a summary of the present-day (1980 to 2009) averages for the temperature variables, based on observed data. This table serves as a comparison for the projected temperature discussion in Section 3.2 (see Table 12 for a description of how these variables were developed).

Table 10: Present-day Averages for Temperature Variables, based on Observed Data
Temperature Event 1980-2009 Temperature Event 1980-2009
Number of days above 95°F 10 days Number of consecutive days above 95°F 4 days
Number of days above 100°F 0.6 days Number of consecutive days above 100°F 0.4 days
Number of days above 105°F 0 days Number of consecutive days above 105°F 0 days
Number of days above 110°F 0 days Number of consecutive days above 110°F 0 days
50th percentile of the hottest day 96.8°F Hottest week of the year 94.4°F
95th percentile of the hottest day 101.3°F Warmest four days in summer 100.8°F
Maximum hottest day 102.8°F    

3.2. Projected Temperature

Climate projections of temperature were statistically downscaled from a number of models and analyzed to project how annual, seasonal, and monthly-average temperature conditions; specific temperature thresholds; and extreme conditions relevant to Mobile, Alabama, may change in the future.

3.2.1. Methodology

This section describes the methodology used to estimate projected temperature changes using daily statistically downscaled climate data from up to ten climate models under three emission scenarios. Additional detail is provided in Appendix C.3.

Climate Models and Emission Scenarios

To capture a range of possible futures, a low emission scenario (B1), a moderately-high emission scenario (A2), and a high emission scenario (A1FI) were selected for a scenario-based analysis of future climate. A summary of the downscaled climate data is provided here (see Hayhoe and Stoner (2012) for a detailed description of the downscaling methodology and validation results).16

Projections of daily temperature for Mobile were statistically downscaled from climate model data housed in the World Climate Research Program (WCRP) Coupled Model Intercomparison Project (CMIP3) multi-model data set.17 Climate models in the WCRP CMIP3 database were selected based on the following three criteria:

Table 11: The Climate Models Used in This Study (Adapted from Hayhoe and Stoner (2012))
Climate Model Name Origin Atmospheric Resolution (horizontal and vertical) Climate Sensitivity (°C) Emission Scenario
B1 A2 A1FI
BCCR-BCM2.0 Bjerknes Centre for Climate Research, Norway 1.9° x 1.9°
31 levels
N/A  
CCSM3 National Center for Atmospheric Research, USA 1.4° x 1.4°
26 levels
2.7
CGCM3 (T47) Canadian Centre for Climate Modeling and Analysis, Canada 2.8° x 2.8°
31 levels
3.4  
CGCM3 (T63) 1.9° x 1.9°
31 levels
3.4  
CNRM-CM3

Météo-France/Centre
National de Recherches
Météorologiques, France

1.9° x 1.9°
45 levels
N/A  
ECHAM5/MPI-OM Max Planck Institute for Meteorology, Germany 1.9° x 1.9°
31 levels
3.4  
GFDL-CM2.0 U.S. Department of Commerce/ National Oceanic and Atmospheric Administration (NOAA)/Geophysical Fluid Dynamics Laboratory (GFDL), USA 2.0° x 2.5°
24 levels
2.9  
GFDL-CM2.1 2.0° x 2.5°
24 levels
3.4
PCM National Center for Atmospheric Research, USA 2.8° x 2.8°
26 levels
2.1
UKMO-Hadcm3 Hadley Centre for Climate Prediction and Research/Met Office, UK 2.5° x 3.75°
19 levels
3.3

To account for local influences, the large-scale climate data was downscaled to the locations of individual local observation stations (see Figure 4).21 The Asynchronous Regional Regression Model (ARRM) method of statistical downscaling was used because it is capable of downscaling at daily timescales.2 To learn more about the downscaling methodology, see Appendix C.2.5 and Hayhoe and Stoner (2012).

Temperature Averages and Events

Table 12 provides a list of the temperature-related weather hazards and climatic averages that are investigated in this study. All of the variables can be directly estimated using the daily downscaled temperature data. These average changes and events were identified and deemed useful by:

Only some of these variables provide robust, quantitative results appropriate for quantitatively-based decisions. In Table 12, asterisks denote the variables and percentiles that do not provide robust quantitative results (per communication with Dr. Hayhoe) and their use should be limited to qualitatively informing the impact assessment. In general, these asterisks denote cases where extreme events are based on a small sample size (e.g., 5th percentile of high daily maximum temperature is based on only 30 data points). In contrast, temperature projections based on a large sample size are considered robust (e.g., average annual temperature is based on 365 days of data per year and averaged across 30 years for a total of more than 10,000 data points).

Table 12: Temperature Variables Developed for this Study
Variable Transportation
Mode
Methodology
Annual, seasonal, and monthly average minimum, maximum, and mean temperature for each 30-year time period Airports (runway design) For each 30-year period, the daily minimum, maximum, and mean temperature corresponding to each month, season, or year were averaged for each station location, climate model, and emission scenario. Then, the 30-year average was determined for each station location, climate model, and emission scenario. Averages and standard deviations were then calculated across climate models for each station location and emission scenario. For purposes of discussion, the results were averaged across station locations to produce an average for the Mobile region.
5th*, 50th and 95th*percentile of high daily maximum temperature and the warmest day of the year for each 30-year time period Rail (AREMA rail design, buildings) For each 30-year period, the daily maximum temperature for each year was identified. This resulted in a total of 30 data points in each time period for each climate model, station location, and emission scenario. The mean, 50th, and 95th percentile levels were estimated from this set of 30 data points applying a quantile distribution and then averaged across climate models for each station location and emission scenario. The warmest day in summer for the 30-year period was estimated in the same way. For purposes of discussion, the results were averaged across station locations to produce an average for the Mobile region.
Seasonal and annual number of days and maximum consecutive days of maximum temperatures at or above 95°F, 100°F, 105°F, and 110°F during each 30-year time period Civil, Geotech, Pavement For each 30-year period, the number of days where the maximum temperature was at or above 95°F, 100°F, 105°F, and 110°F was counted for each year. This resulted in 30 data points in each time period (one for each year), for each climate model, station location, and emission scenario. The 30 data points were averaged to estimate the annual number of days at or above each high temperature threshold for each climate model, station location, and emission scenario. The mean and standard deviation was then determined across the climate models for each station location and emission scenario. The process was repeated to obtain seasonal projections. The maximum consecutive days of high temperature for each threshold was likewise calculated. For purposes of discussion, the results were averaged across station locations to produce an average for the Mobile region.
Mean; 5th*, 25th, 50th, 75th, and 95th* percentile; and minimum value for the average minimum air temperature over four consecutive days in winter and the average maximum temperature over four consecutive days in summer for each 30-year time period Bridge, Rail For each winter in the 30-year period, the average of the minimum air temperature for any four consecutive days was estimated for each climate model projection, emission scenario, and location. The 5th, 25th, 50th, 75th, and 95th percentile; mean; and coldest period across the 30 data points was estimated for each climate model, emission scenario, and location applying a quantile distribution. For each summer in the 30-year period, the average of the maximum air temperature for any four consecutive days was estimated for each year, ultimately providing the 5th, 25th, 50th, 75th, 95th percentile; mean; and hottest period across the 30 data points for each climate model, emission scenario, and location. The average across climate models for each location was determined, and then averaged across station locations to provide an average for the Mobile region.
The average, 1st*, 5th *, 10th, and 50th percentile of the coldest day of the year during each 30-yr time period Multi (pavement design) Using the daily minimum temperatures, the coldest minimum temperature for each year was identified for each climate model, emission scenario, and station location. Across the 30 data points for each time period, the mean, 1st, 5th, 10th, and 50th percentile was calculated by applying a quantile distribution for each climate model, emission scenario, and station location. The average across climate models for each location was determined, and then averaged across station locations to provide an average for the Mobile region.
Maximum 7-day average air temperature per year with the % probability of occurrence during each 30-yr period (mean, 50th, 90th, 95th*, 99th*percentile) for each 30-yr time period Multi (pavement design - asphalt) Using the daily maximum temperature, the maximum 7-day average temperature for each year was determined. This produced a total of 30 data points in each time period, for each climate model, emission scenario, and station location. Across the 30 data points, the mean, 50th, 90th, 95th, and 99th percentile was estimated by applying a quantile distribution for each climate model, emission scenario, and station location. The average across climate models for each location was determined, and then averaged across station locations to provide an average for the Mobile region.

The statistically downscaled climate model results were provided for each emission scenario, for each of the five station locations, and for the baseline and projected time frames (1980 to 2009, 2010 to 2039, 2040 to 2069, and 2070 to 2099). For each station location, the results were then averaged across the statistically downscaled climate models to produce a climate model ensemble average (the average projection across all climate models for a given emission scenario and time period).23 The projected change for each temperature average, threshold, and extreme was calculated by comparing the climate model ensemble average projections to the climate model ensemble average baseline ("simulated baseline") of 1980 to 2009.2

Identification of Statistically Significant Climate Projections

This study produced an enormous amount of climate projections across station locations, emission scenarios, and time periods. For a complete database of the climate projections by emission scenario, location, and time period, see Appendix E.1.

To focus the study on climate projections that represent a statistically significant projected change from the simulated baseline conditions, a statistical test (a paired t-test) was used to identify significant (p<0.05) changes, i.e., climate projections that are statistically different from simulations of the baseline climate (1980 to 2009). See Appendix C.3.2 for a description of the paired t-test.

For the purposes of this study, only the variables that demonstrate a statistically significant change at all five station locations are considered to demonstrate statistically significant differences for the region.

3.2.2. Key Findings

This section presents the results of the projected temperature analysis, including a discussion of:

Temperature is projected to increase over time as greenhouse gas concentrations in the Earth's atmosphere continue to increase. The farther out in time, the greater the amount of temperature increase. Overall, the amount of temperature increase is directly proportional to the increase in emissions—that is, the high (A1FI) emission scenario is associated with greater overall temperature increases than the low (B1) emission scenario.

The increase in seasonal and monthly means is more variable across the emission scenarios. For example, under the low (B1) and moderately-high (A2) emission scenarios, seasonal average temperatures are projected to increase the most in the fall season, with monthly average temperatures increasing the most in October. Under the high (A1FI) emission scenario, seasonal average temperatures are projected to increase the most in the summer season, with monthly average temperatures still increasing the most in October. As emissions increase there may be a tendency for peak warming to shift from fall to summer seasons.

Explanation of Box Plots

Throughout this report, the projected changes in environmental variables are illustrated by box plots. Each box on the plot shows the mean (represented by the line separating the two types of shading) and variability (represented by the box height) of climate projections for each time period and emission scenario, averaged across all five stations and the climate model ensemble. The 'simulated baseline' is the average temperature simulated from 1980 to 2009, as modeled by all downscaled climate models and averaged across emission scenarios and station locations.

Tables including more detail about the projected changes are available in Appendix C.5.

Average Annual Temperatures

Key Findings for Average Annual Temperatures

  • Average maximum, minimum, and mean temperatures are projected to increase significantly in each time period under all emission scenarios.
  • Average mean temperatures increase steadily with each 30-year time period by approximately 1oF (0.6°C), 2oF (1°C), and 3oF (2°C) for the low (B1), moderately-high (A2), and high (A1FI) emission scenarios, respectively.
  • Minimum temperatures are projected to increase more than maximum temperatures.

Average annual maximum, minimum, and mean temperatures averaged across each of the time periods are projected to increase significantly (p<0.05) for all emission scenarios. Figure 10 illustrates how average annual mean temperatures are projected to increase over time in the Mobile region as a function of emission scenario and time period.

Figure 10: Projected Average Mean Temperature (°F)

This figure shows projections of average daily mean temperature for the B1, A2, and A1FI emission scenarios in 2010-2039, 2040-2069, and 2070-2099. The figure also shows the average modeled baseline projections for 1980-2009. Each bar shows the multi-model mean for that emission scenario and the range of plus/minus one standard deviation. All three scenarios show similar projections of approximately 69 degrees in the near-term. In the 2040-2069 and 2070-2099 time frames, the B1 scenario projects the lowest average temperatures, followed by the A2 then A1FI scenarios.

Figure 10 also illustrates that the variability of the temperature projections increases with time (variability is indicated by the range of projections across both climate models and station locations).25 This figure shows that though temperature increases in the near-term are similar under all three emission scenarios, a distinct pattern based on emission scenario emerges by mid-century and end-of-century. By mid-century, average annual mean temperatures are projected to increase 2.4°F (1.3°C), 3.5°F (1.9°C), and 4.6°F (2.6°C) above modeled baseline for the low (B1), moderately-high (A2), and high (A1FI) emission scenarios, respectively. By end-of-century, average annual mean temperatures are projected to increase by 3.2°F (1.8°C), 6.6°F (3.7°C), and 7.7°F (4.3°C) above modeled baseline, for the low (B1), moderately-high (A2), and high (A1FI) emission scenarios, respectively. A noticeable pattern suggests average annual mean temperatures steadily increase with each 30-year time period by approximately 1°F (0.6°C), 2°F (1°C), and 3°F (2°C) for the low (B1), moderately-high (A2), and high (A1FI) emission scenarios, respectively.

The average annual mean temperature observed over the past 30 years (1980-2009) in the Mobile region was about 67°F (19°C). By the end-of-century, this may increase to 70.5°F (21.4°C) under the low (B1) emission scenario, 73.8°F (23.2°C) under the moderately-high (A2) emission scenario, and 74.8°F (23.8°C) under the high (A1FI) emission scenario.

Average annual maximum temperatures are projected to increase from 78°F (26°C) at baseline to as high as 84°F (29°C) by the end-of-century under the high (A1FI) emission scenario.

Figure 11 and Figure 12 show that minimum temperatures are projected to increase more than maximum temperatures. This suggests that the diurnal temperature range will reduce with time.

Figure 11: Projected Average Minimum Temperature (°F)

This figure shows a bar graph with projections of average daily minimum temperature for the B1, A2, and A1FI emission scenarios in 2010-2039, 2040-2069, and 2070-2099. The figure also shows the average modeled baseline projections for 1980-2009 (about 57 degrees Fahrenheit). Each projection bar shows the multi-model mean for that emission scenario and the range of one standard deviation above and below the mean. The figure shows that temperatures are projected increase both with time and with emissions. In the 2010-2039 time frame, all three scenarios show similar projections of approximately 58 degrees. In the 2040-2069 and 2070-2099 time frames, the B1 scenario projects the lowest minimum temperatures, followed by the A2 then A1FI scenarios. The highest projection of average minimum temperature is 66 degrees, under the A1FI scenario for 2070-2099. The uncertainty associated with the projections also increases in each time period, and is largest in the A2 scenario.

Figure 12: Projected Average Maximum Temperature (°F)

This figure shows a bar graph with projections of average daily maximum temperature for the B1, A2, and A1FI emission scenarios in 2010-2039, 2040-2069, and 2070-2099. The figure also shows the average modeled baseline projections for 1980-2009 (about 78 degrees Fahrenheit). Each projection bar shows the multi-model mean for that emission scenario and the range of plus/minus one standard deviation. In the 2010-2039 time frame, all three scenarios show similar projections of approximately 80 degrees. In the 2040-2069 and 2070-2099 time frames, the B1 scenario projects the lowest minimum temperatures, followed by the A2 then A1FI scenarios. The range of projections also increases in each time period. The highest projection of average minimum temperature is 84 degrees, under the A1FI scenario for 2070-2099.

Key findings for average annual temperature increases relative to simulated baseline (1980-2009) are as follows:

Season Definitions

Winter = December, January, February
Spring = March, April, May
Summer = June, July, August
Fall = September, October, November

Average Seasonal and Monthly Temperatures

Key Findings for Average Seasonal and Monthly Temperatures

Average seasonal and monthly mean temperatures are also projected to increase significantly for all emission scenarios and time periods. Projected changes in seasonal and monthly average mean temperatures are driven largely by the projected changes in average minimum temperatures.

Under the low (B1) and moderately-high (A2) emission scenarios, seasonal average temperatures are projected to increase the most in the fall season. Monthly average temperatures are projected to increase the most in October.

Under the high (A1FI) emission scenario, seasonal average temperatures are projected to increase the most in the summer season. Monthly average temperatures are still projected to increase the most in October.

Projections suggest that the average seasonal and monthly diurnal temperature range may decrease, particularly for temperature projections associated with the moderately-high (A2) and high (A1FI) emission scenarios.

Key findings for seasonal and monthly temperature increases relative to simulated baseline (1980 to 2009) are as follows:

Figure 13: Change in Average Seasonal Mean Temperature, Averaged Across Climate Models and Station Locations for Mid-Century (2040-2069) Relative to Baseline (1980-2009)

This figure shows the projected change in average seasonal temperatures from 1980-2009 to mid-century (2040-2069) for each of the three emission scenarios. For each season, the projections under A1FI scenario show the largest change while projections under the B1 scenario show the smallest change. For winter, the figure shows projected changes of 2.2, 3.0, and 3.3 degrees; for spring, changes of 2.1, 3.0, and 4.7 degrees; for summer, changes of 2.4, 3.8, and 5.5 degrees; and for fall, changes of 2.9, 4.2, and 5.1 degrees for the B1, A2, and A1FI scenarios, respectively.

Figure 14: Projected Change in Average Monthly Mean Temperature, Averaged Across Climate Models and Station Locations for Mid-Century (2040-2069) Relative to Baseline (1980-2009)

This figure shows the projected change in average monthly temperatures from 1980-2009 to mid-century (2040-2069) for the B1, A2, and A1FI emission scenarios. For all months except January and December, the projections under A1FI scenario show the largest change while projections under the B1 scenario show the smallest change. B1 scenario projections do not vary significantly by month and show an increase in temperatures of approximately 2 degrees. The largest change for B1 is in November. A2 scenario projected changes are lowest in February (2.1 degrees) and increase steadily through October (4.75 degrees). A1FI scenario changes range from 3 to 5.8 degrees, with the largest changes in the summer and early fall months.

Figure 15: Change in Average Seasonal Mean Temperature, Averaged Across Climate Models and Station Locations for End-of-Century (2070-2099) Relative to Baseline (1980-2009)

This figure shows the projected change in average seasonal temperatures from 1980-2009 to end-of-century (2070-2099) for each of the three emission scenarios. For each season except winter, the projections under A1FI scenario show the largest change while projections under the B1 scenario show the smallest change. For winter, the figure shows projected changes of 2.9, 5.7, and 5.7 degrees; for spring, changes of 2.8, 6.0, and 7.8 degrees; for summer, changes of 3.0, 6.9, and 9.0 degrees; and for fall, changes of 4.2, 7.9, and 8.5 degrees for the B1, A2, and A1FI scenarios, respectively.

Figure 16: Projected Change in Average Monthly Mean Temperature, Averaged Across Climate Models and Station Locations for End-of-Century (2070-2099) Relative to Baseline (1980-2009)

This figure shows the projected change in average monthly temperatures from 1980-2009 to mid-century (2040-2069) for the B1, A2, and A1FI emission scenarios. For all months except January and November, the projections under A1FI scenario show the largest change while projections under the B1 scenario show the smallest change. B1 scenario projections do not vary significantly by month but show the largest changes in the fall months, of approximately 4.2 degrees. A2 scenario projected changes are lowest in February (4.9 degrees) and increase steadily through October (8.9 degrees). A1FI scenario changes range from 5.2 to 9.3 degrees, with the largest changes from May through October.

Figure 17: Projected Change in Average Monthly Minimum and Maximum Temperatures, Averaged Across Climate Models and Station Locations for End-of-Century (2070-2099) Relative to Baseline (1980-2009)

This figure shows the projected changes from baseline to end-of-century in minimum and maximum temperatures. The graph shows that for all emission scenarios, minimum temperatures are projected to increase more than maximum temperatures. This gap is largest in the A1FI scenario and in the months of May through October.

A common theme across these projections is that the range between minimum and maximum temperatures is projected to decrease over time (see Figure 17). A decrease in the range of daily temperatures may benefit pavement and other infrastructure, but the level of projected temperatures under the future scenarios implies that there may be less cooling relief overnight.

Extreme Temperature Events

Key Findings for Extreme Temperature Events

The discussion of extreme temperature events are divided into two sections: extreme heat and extreme cold. See Table 10 for the present-day (1980 to 2009) averages for the temperature variables, based on observed data.

Extreme Heat

The discussion of extreme heat is further divided into two sections: hot days and hot periods.
Hot days
This section presents the projected change for the following variables:

These variables were derived from maximum temperature and averaged across the downscaled climate model and station location for each emission scenario and future time period (see Table 12 for a description of the how each of these variables were calculated).

From 1980 to 2009, the hottest temperature of the year in the Mobile region averaged nearly 103°F (39°C). This high temperature is very unusual, as Mobile only experiences an average of 9 to 10 days per year above 95°F (35°C) and less than one day per year above 100°F (38°C). The duration of heat events (i.e., number of consecutive days when the daily maximum temperature is above 95°F (35°C)) averaged almost four days per year.

The number of days above 95°F (35°C) and 100°F (38°C) is projected to increase significantly. However, there was no significant projected change in the number of days above 105°F (41°C) or 110°F (43°C) (currently zero – see Table 10). Only the end-of-century moderately-high (A2) and high (A1FI) emission scenarios project an average 1 to 2 days of temperatures above 105°F (41°C). Figure 18 and Figure 19 show the dramatic projected increases in the number of hot days.

By mid-century, the number of days above 95°F (35°C) is projected to increase by two weeks under the low (B1) emission scenario and by a month or more under the moderately-high (A2) and high (A1FI) emission scenarios. This represents an approximately four-fold increase over baseline.

By end-of-century, the number of days above 95°F (35°C) is projected to increase by three weeks under the low (B1) emission scenario and by two months or more under the moderately-high (A2) and high (A1FI) emission scenarios. This represents a nine-fold increase in the number of days over 95°F (35°C) under the A1FI scenario by end-of-century.

Figure 18: Projected Number of Days per Year above 95°F

This figure shows a bar graph with projections of the number of days above 95 degrees for the B1, A2, and A1FI emission scenarios in 2010-2039, 2040-2069, and 2070-2099. The figure also shows the average modeled baseline projections for 1980-2009 (about 8 days). Each projection bar shows the multi-model mean for that emission scenario and the range of one standard deviation above and below the mean. The figure shows that temperatures are projected increase both with time and with emissions. In the 2010-2039 time frame, all three scenarios show similar projections of approximately 17 days. In the 2040-2069 and 2070-2099 time frames, the B1 scenario projects the fewest days above 95, followed by the A2 then A1FI scenarios. The highest projection of number of days above 95 is 84 days, under the A1FI scenario for 2070-2099. The uncertainty associated with the projections also increases in each time period, and is largest in the A2 scenario.

Figure 19: Projected Number of Days per Year above 100°F

This figure shows a bar graph with projections of the number of days above 100 degrees for the B1, A2, and A1FI emission scenarios in 2010-2039, 2040-2069, and 2070-2099. The figure also shows the average modeled baseline projections for 1980-2009 (about 0.2 days). Each projection bar shows the multi-model mean for that emission scenario and the range of one standard deviation above and below the mean. The figure shows that temperatures are projected increase both with time and with emissions. In the 2010-2039 time frame, all three scenarios show similar projections of approximately 1 day above 100. In the 2040-2069 and 2070-2099 time frames, the B1 scenario projects the fewest days above 100, followed by the A2 then A1FI scenarios. The highest projection of number of days above 100 is 20 days, under the A1FI scenario for 2070-2099. The uncertainty associated with the projections also increases in each time period, and is largest in the A2 scenario.

Figure 20 and Figure 21 show the increase in the maximum number of consecutive days over 95°F (35°C) and 100°F (38°C). These figures suggest that heat events will last longer in the future.

Figure 20: Projections of the Longest Number of Consecutive Days per Year above 95°F

This figure shows a bar graph with projections of the longest number of consecutive days above 95 degrees for the B1, A2, and A1FI emission scenarios in 2010-2039, 2040-2069, and 2070-2099. The figure also shows the average modeled baseline projections for 1980-2009 (about 4 days). Each projection bar shows the multi-model mean for that emission scenario and the range of one standard deviation above and below the mean. The figure shows that temperatures are projected increase both with time and with emissions. In the 2010-2039 time frame, all three scenarios show similar projections of approximately 10 days. In the 2040-2069 and 2070-2099 time frames, the B1 scenario projects the fewest consecutive days above 95, followed by the A2 then A1FI scenarios. The highest projection of consecutive days above 95 is 39 days, under the A1FI scenario for 2070-2099. The uncertainty associated with the projections also increases in each time period, and is largest in the A2 scenario.

Figure 21: Projections of the Longest Number of Consecutive Days per Year above 100°F

This figure shows a bar graph with projections of the longest number of consecutive days above 100 degrees for the B1, A2, and A1FI emission scenarios in 2010-2039, 2040-2069, and 2070-2099. The figure also shows the average modeled baseline projections for 1980-2009 (about 0.2 days). Each projection bar shows the multi-model mean for that emission scenario and the range of one standard deviation above and below the mean. The figure shows that temperatures are projected increase both with time and with emissions. In the 2010-2039 time frame, all three scenarios show similar projections of approximately 1 day. In the 2040-2069 and 2070-2099 time frames, the B1 scenario projects the fewest consecutive days above 100, followed by the A2 then A1FI scenarios. The highest projection of consecutive days above 100 degrees is 9 days, under the A1FI scenario for 2070-2099. The uncertainty associated with the projections also increases in each time period, and is largest in the A2 scenario.

Figure 22 and Figure 23 illustrate percentiles of temperature for the hottest day of the year by time period and emission scenario. As with other temperature metrics, the temperature of the hottest day of the year is projected to increase with time and emissions. Variability in projections across climate models and observation stations also increases over time. The maximum hottest temperature projected by end-of-century is 109.3°F (42.9°C), under the A1FI emission scenario.

Figure 22: Hottest Day of the Year, Mid-Century (2040-2069)

This figure shows a bar graph with projections of the hottest day of the year by mid-century under the B1, A2, and A1FI emission scenarios. The projections are shown for the 50th percentile of projected high temperatures, 95th percentile of projected high temperatures, and highest projected temperature. Each projection bar shows the multi-model mean for that emission scenario and the range of one standard deviation above and below the mean. The figure shows that temperatures are projected increase with emissions, and that there is little difference in trends between the 50th percentile, 95th percentile, and maximum projections. The projected hottest day of the year is 105 degrees under the B1 scenario and 108 degrees under the A1FI scenario. The projected 50th percentile ranges between 100 degrees under B1 to 102 degrees under A1FI.

Figure 23: Hottest Day of the Year, End-of-Century (2070-2099)

This figure shows a bar graph with projections of the hottest day of the year by the end of the century under the B1, A2, and A1FI emission scenarios. The projections are shown for the 50th percentile of projected high temperatures, 95th percentile of projected high temperatures, and highest projected temperature. Each projection bar shows the multi-model mean for that emission scenario and the range of one standard deviation above and below the mean. The figure shows that temperatures are projected increase with emissions, and that there is little difference in trends between the 50th percentile, 95th percentile, and maximum projections. The projected hottest day of the year is 106 degrees under the B1 scenario and 111 degrees under the A1FI scenario. The projected 50th percentile ranges between 102 degrees under B1 to 105 degrees under A1FI. The A2 scenario shows the largest uncertainty in projections.

Key findings for projected increases in the number of days above a given threshold relative to simulated baseline (1980-2009) are as follows (see Table 10 for present-day conditions observed in the Mobile region from 1980 to 2009):

The figures above demonstrate the spread of estimates across the climate models for each emission scenario. The greatest spread is associated with the moderately-high (A2) emission scenario, which is determined from ten climate models. The spread associated with the high (A1FI) emission scenario is artificially small in comparison, as it is only determined from four climate models.
Hot Periods
This section presents the projected change for the following variables:

These variables were derived from maximum temperature and averaged across the downscaled climate model and station location for each emission scenario and future time period (see Table 12 for a description of the how each of these variables were calculated).

Projected temperature was also analyzed for the hottest week of the year, the warmest four days in summer, and the warmest summer in 30 years. Figure 24 shows probability curves of the warmest summer simulated for each time period and for each emission scenario. It illustrates how temperatures are projected to shift (e.g., the area under the curve illustrates how likely the temperature is projected to occur).26 The curves demonstrate how the average summer temperature may change, reflected by the top of the bell-shaped curve. The curves also demonstrate how the extremely cold or extremely warm summer average temperature may change, reflected by the tails of the curve.

Key findings for these projections relative to simulated baseline (1980-2009) are as follows:

Figure 24 illustrates the change in the distribution of the warmest four days in summer over time for each emission scenario. The tails of the projections represent the coldest and hottest warmest four days in summer for each time period and, given the uncertainty in the tails, should be considered representative of possible extreme values; meanwhile, the peak of the curve represents the average warmest four days in summer. Projections under the low (B1) emission scenario suggest a shift to warmer conditions. Projections associated with the moderately-high (A2) and high (A1FI) emission scenarios suggest that the probability of the warmest four days of summer to become hotter over the 30-year period will increase dramatically over time compared to baseline conditions.

Figure 24: Warmest 4 Days in Summer, Projections by Emission Scenario Averaged Across Climate Models and Station Locations Relative to Model Baseline (1980-2009)

This figure shows the probability of the temperature of the warmest four days in summer for each time period. The figure shows one bell-shaped curve for each time period. Under all scenarios, summer temperatures are projected to become warmer over time, with the peak of the bell curve shifting to the right. Also under all scenarios, there is about a 15 percent probability that the warmest four days in the summer will average 90 degrees. Under the B1 scenario, is a 16 percent probability that temperatures will be 93 degrees by end of century. Under the A2 scenario, there is a 15 percent probability that those days will average about 96 degrees by 2070-2099. Under the A1FI scenario, there is a 15 percent probability that the warmest four days of summer will average 97 degrees. This figure shows the probability of the temperature of the warmest four days in summer for each time period. The figure shows one bell-shaped curve for each time period. Under all scenarios, summer temperatures are projected to become warmer over time, with the peak of the bell curve shifting to the right. Also under all scenarios, there is about a 15 percent probability that the warmest four days in the summer will average 90 degrees. Under the B1 scenario, is a 16 percent probability that temperatures will be 93 degrees by end of century. Under the A2 scenario, there is a 15 percent probability that those days will average about 96 degrees by 2070-2099. Under the A1FI scenario, there is a 15 percent probability that the warmest four days of summer will average 97 degrees. This figure shows the probability of the temperature of the warmest four days in summer for each time period. The figure shows one bell-shaped curve for each time period. Under all scenarios, summer temperatures are projected to become warmer over time, with the peak of the bell curve shifting to the right. Also under all scenarios, there is about a 15 percent probability that the warmest four days in the summer will average 90 degrees. Under the B1 scenario, is a 16 percent probability that temperatures will be 93 degrees by end of century. Under the A2 scenario, there is a 15 percent probability that those days will average about 96 degrees by 2070-2099. Under the A1FI scenario, there is a 15 percent probability that the warmest four days of summer will average 97 degrees.

Extreme Cold

This section presents the projected change for the following variables:

These variables were derived from minimum temperature and averaged across the downscaled climate model and station location for each emission scenario and future time period (see Table 12 for a description of the how each of these variables were calculated).

Along with increases in the temperature, frequency, and duration of heat events, the Mobile region is projected to experience significant decreases in the occurrence of cold temperatures.

Winter temperatures are projected to warm, such that the average coldest four days in winter are projected to be nearly 3°F (2°C) warmer by the end of the century under the low (B1) emission scenario, and nearly 6°F (3°C) warmer under the moderately-high (A2) and high (A1FI) emission scenarios.

The coldest day of the year in the baseline period averaged 19°F (-7°C) in the Mobile region. The coldest day is projected to warm significantly over time, with the average coldest day ranging from 21 to 22°F (-6 to -5.6°C) at mid-century (see Figure 25), and ranging from 21 to 25°F (-6 to -3.9°C) at end-of-century (see Figure 26), depending on emission scenarios. Projections of the coldest day of the year suggest that the extreme cold day in a 30-year time period will warm substantially more than the average cold day in the same time period. The climate model projections, however, show high variability when projecting the lowest temperature of the year.

Key findings for extreme cold relative to simulated baseline (1980-2009) are as follows:

Figure 25: Coldest Day of the Year (°F), Mid-Century (2040-2069)

This figure shows a bar graph with projections of the coldest day of the year by mid-century under the B1, A2, and A1FI emission scenarios. The projections are shown for the 1st, 5th, 10th, and 50th percentile of projected low temperatures. Each projection bar shows the multi-model mean for that emission scenario and the range of one standard deviation above and below the mean. The figure shows that temperatures are projected increase with emissions, and that there is little difference in trends between the different percentile projections--except for the 1st percentile, coldest day temperatures are projected to be lowest under B1 and warmest under A1FI. The uncertainty associated with the projections is largest under the A1FI scenario. The projected 1st percentile of the coldest day of the year is 11 degrees under the B1 scenario and 14 degrees under the A1FI scenario. The projected 50th percentile ranges between 22 degrees under B1 to 23 degrees under A1FI.

Figure 26: Coldest Day of the Year (°F), End-of-Century (2069-2099)

This figure shows a bar graph with projections of the coldest day of the year by the end of the century under the B1, A2, and A1FI emission scenarios. The projections are shown for the 1st, 5th, 10th, and 50th percentile of projected low temperatures. Each projection bar shows the multi-model mean for that emission scenario and the range of one standard deviation above and below the mean. The figure shows that temperatures are projected increase with emissions, and that there is little difference in trends between the different percentile projections-- coldest day temperatures are projected to be lowest under B1 and warmest under A1FI. The projected 1st percentile of the coldest day of the year is 10 degrees under the B1 scenario and 15 degrees under the A1FI scenario. The projected 50th percentile ranges between 22 degrees under B1 to 25 degrees under A1FI.

Figure 27 shows how the coldest four days in winter in each time period is projected to become warmer under all emission scenarios. Given the uncertainty in the tails of the projections, the shifts in these extremes should be used cautiously when informing transportation planning. The mean temperature of the coldest four days in winter increases under all emission scenarios and time periods. Unlike the projections of the warmest four days in summer, the shape of the curves tends to remain somewhat consistent over time. The coldest four days in winter observed from 1980 to 2009 averaged 41°F (5°C). By end-of-century, the coldest four days in winter is projected to average 44 to 47°F (7 to 8°C), depending on emission scenario.

Figure 27: Coldest 4 Days in Winter, Projections by Emission Scenario Averaged Across Climate Models and Station Locations Relative to Baseline (1980-2009)

This figure shows the probability of the temperature of the coldest four days in winter for each time period. The figure shows one bell-shaped curve for each time period. Under all scenarios, winter temperatures are projected to become warmer over time, with the peak of the bell curve shifting to the right. Also under all scenarios, there is about a 6 percent probability that the coldest four days in the summer will average 42 degrees. Under the B1 scenario, is a 6 percent probability that temperatures will be 44 degrees by end of century. Section of figure showing projections under A2. Under the A2 scenario, there is a 6 percent probability that those days will average about 46.5 degrees by 2070-2099. Section of figure showing projections under A1FI. Under the A1FI scenario, there is a 6 percent probability that the coldest four days of winter will average 46.5 degrees.

3.3. Implications for Transportation

These projected changes in temperature have some notable implications for transportation infrastructure and services. In general, higher temperatures result in more rapid deterioration of pavements that could require changes in repair and maintenance schedules (although in the longer term, newer and more durable pavement designs could reduce this impact). In addition, longer growing seasons due to longer periods of warmer temperatures could require more attention to mowing in rights of way, thus affecting maintenance budgets. An increase in the duration and frequency of extreme temperature events can result in increased buckling of rail and rutting and shoving of pavement. These impacts could be exacerbated by reduced potential for cooling relief overnight for pavement and other infrastructure. Excessive heat can contribute to equipment failures and more frequent vehicle breakdowns. Energy requirements for air conditioning of buildings, equipment, transit facilities, and freight are likely to increase. Ports, in particular, may see increases in energy costs to meet air conditioning and refrigeration requirements.

Extreme heat events also have health and safety implications for transportation agency personnel. In particular, maintenance and construction schedules may need to be adjusted to avoid health risks to workers. Further, the costs of ensuring the comfort and safety of passengers – particularly of train and bus travelers – are likely to increase.

The implications of the temperature findings detailed in this report on transportation assets and services in Mobile will be investigated in the next task of this study (Task 3: Vulnerability Screen and Assessment).


1 NRC, 2008; USCCSP, 2008a

2 NRC, 2008; USCCSP, 2008a

3 NRC, 2008

4 Historical data was provided by Dr. Katharine Hayhoe of Texas Tech, who also conducted the temperature trend analysis.

5 Daily downscaled projections were provided by Dr. Katharine Hayhoe of Texas Tech,

6 Dr. Katharine Hayhoe of Texas Tech provided the historical data (see Hayhoe and Stoner 2012 for a discussion of data quality and additional data filtering).

7 The GHCN dataset was sufficient as it was important to: (1) have daily data for event analysis, and (2) to use a consistent set of data when comparing against the climate projections later in this report. Had these two factors not been important, the homogeneity adjusted data in the USHCN would have been considered. The USHCN data provide monthly and seasonal data (http://www.ncdc.noaa.gov/oa/climate/research/ushcn/). http://www.ncdc.noaa.gov/oa/climate/research/ushcn/), as recommended per communication with Dr. Tom Peterson of NOAA, with careful consideration in treatment of the data gaps existing in the record. Hence, the USHCN data set has no data gaps. To determine what impact the daily data gaps that exist in the GHCN data may have had, monthly and seasonal averages using the GHCN data set were compared against the USHCN data set for the one location in common (Fairhope). This comparison indicated no noticeable differences, suggesting the daily data gaps in Fairhope observational record have minimal impact on the monthly and seasonal averages.

8 The estimates of mean annual temperature were calculated by Dr. Katharine Hayhoe.

9 The results of the Mann Kendall trend analysis for the historical record were provided by Dr. Katharine Hayhoe.

10 Statistical significance at p<0.10 represents a confidence interval level of 90%. The trend for each station was considered when determining if a regional trend was noted (i.e., all five stations needed to agree). The "p-value" is a test statistic that suggests whether a significant change in temperature has occurred over the time period considered. A p –value of below 10% is less than a one-in-10 chance that the difference observed over time is due to chance. The smaller the p-value, the less likely it is that the trend observed is due to chance (e.g., a p-value of 5% is a one-in-20 chance that the trend observed is due to chance).

11 The averaging for the region provides a means to effectively communicate local changes and provides a larger sample size to support the findings than based on simply one location alone. Averaging across all stations and years for regional average annual temperatures may skew the data towards the locations with the most complete and longest data record. A check was conducted, and it was determined that an artificial trend did not appear that could be associated with any one station.

12 Temperature averages provided in Table 3, 4, and 5 for the 1980-2009 time period are based on the observational data. This time period is consistent with the baseline time period used later in the temperature projections piece (e.g., the baseline 1980-2009 climate model simulations).

13 The 1961 to 2010 analysis is based on full data except as noted previously and is for a partially complete 2010 year (does not include October, November, and December).

14 Hayhoe and Stoner (2012) tested significance using the non-parametric Mann-Kendall analysis. See Hayhoe and Stoner (2012) for a full discussion of the Mann-Kendall analysis methodology.

15 The 10-year moving average is provided for each figure. As the five stations provide varying data records, the moving average from 1920 to 1960 is based on Bay-Minette and Fairhope station data, and the moving average after 1960 is an average across all five station data. An additional screening was applied to the observation data to remove any years where the hottest day of the year was observed to be more than three standard deviations from the long-term average.

16 The statistically downscaled projections were provided by Dr. Katharine Hayhoe and Dr. Anne Stoner through an interagency agreement between FHWA and USGS (see Hayhoe and Stoner (2012) for a detailed analysis and summary of the methodology and results). The WCRP CMIP3 data set can be found at http://esg.llnl.gov:8080/index.jsp. For purposes of this study, climate model is used synonymously with global climate model (GCMs).

17 The WCRP CMIP3 data set was used to inform the findings of the Intergovernmental Panel on Climate Change (IPCC) Fourth Assessment Report (AR4). This database will be replaced with the WCRP CMIP5 data set that is currently being developed by the climate community and whose findings will be summarized in the IPCC Fifth Assessment Report (AR5) scheduled to be published in September 2013. The methodologies and tools developed in this report are transferable to the new data set.

18 Note that the models which simulate the hottest temperature response to a doubling of CO2 concentrations (i.e., highest climate sensitivity) are not included in this assessment.

19 Hayhoe and Stoner, 2012.

20 All ten models had projections for the B1 and A2 emission scenarios. Only four of the models as noted in Table 11 had projections for the A1FI emission scenario.

21 Downscaling of the global climate model simulations was conducted by Dr. Katharine Hayhoe.

22 Hayhoe and Stoner, 2012.

23 Since there are 10 GCMs providing results for the A2 and B1 emission scenarios, the uncertainty estimates include ranges of one standard deviation from the mean based on the set of all relevant climate model simulations. Since only four GCMs provide results for the A1FI emission scenario, the uncertainty estimates are a coarser range of model results described by the minimum and maximum GCM values.

24 The climate model simulations of 1980 to 2009 are similar but not identical to the observations of 1980 to 2009 discussed in section 2.1.1.

25 The trends were found to be consistent across all five stations.

26 These plots were developed by fitting a standard Gaussian distribution using the values provided for the 5%-, 25%-,50%-, 75%-, 90%-, and 95%-percentile for the warmest summer.

Updated: 10/31/2014
HEP Home Planning Environment Real Estate
Federal Highway Administration | 1200 New Jersey Avenue, SE | Washington, DC 20590 | 202-366-4000