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

Temperature and Precipitation Projections for the Mobile Bay Region

Section 4: Changes in Precipitation

Climate change is not just about warmer temperatures; as the planet warms, precipitation patterns are also expected to shift in both space and time. Some seasons may get wetter, while others get drier. The intensity and frequency of heavy rainfalls, as well as the duration of dry periods, may be altered.

This section summarizes observed historical trends and the changes in precipitation and rainfall-related secondary indicators that have been observed and are projected to occur in the Mobile Bay area in response to global change. As in Section 3, historical trends were detected using a Mann-Kendall trend analysis. A "significant" trend is defined here as one with a p-value less than 0.1, and a "consistent" trend is one where multiple stations have the same sign of the Kendall 1. Projected future changes are consistent across all five stations; unless otherwise indicated, plotted values correspond to the average value across the five stations.

Figure 4.1 - Click for details

Figure 4.1 Projected annual average precipitation for the average of the 5 weather stations, as simulated by the average of ten climate models for the B1 (lower), A2 (mid-high) and A1FI (higher) emissions scenarios. Error bars show range of projected values for A1fi and 2s range for A2 & B1 (i.e., one s above and below the mean).

This figure assesses the trend in observed average seasonal and monthly precipitation from 1960-2010 in Mobile.  The Kendall tau shows whether the trend is negative or positive, and the strength of the trend (between -1.0 and +1.0).  For summer the trend is consistent across all five stations, and strong at about 0.7.  For fall and five months of the year, the trend is increasing across four of the five stations, with a value of between about 0.5 and 0.7

Figure 4.2 Data from 1960 to 2010 shows a significant increase in average precipitation at all five stations for one season (summer, dark red) and at four out of five stations for one more season (fall, red) and five months of the year.

Figure 4.2 (winter) - Click for detailsFigure 4.2 (spring) - Click for details

Figure 4.2 (summer) - Click for detailsFigure 4.2 (fall) - Click for details

Figure 4.3 Projected change in seasonal average precipitation averaged across the five Mobile Bay weather stations relative to 1980-2009; as simulated by the average of ten climate models for the B1 (lower), and A2 (mid-high) scenarios and four climate models for the A1FI (higher) emissions scenarios. Error bars show range of projected values for A1Fi and 2s range for A2 & B1 (i.e., one s above and below the mean).

Annual and Seasonal Precipitation

Annual precipitation in the Mobile Bay region averages around 65 inches per year. From 1960 through 2010, no significant changes were observed in this annual average. Over this century, there is some indication, although with significant variability, of a relatively small increase of several inches by mid-century, followed by a return to present-day values under higher emissions scenarios by end-of-century (Fig. 4.1).

Little to no change in annual average precipitation can mask significant changes in the seasonal and monthly distribution of precipitation. From 1960 to 2010, for example, a significant and consistent increase in summer (JJA) precipitation was observed across all five stations. When the definition of consistency was relaxed to require just four out of five stations to show a trend in the same direction, there were significant increases in summer (JJA) and fall (SON) precipitation, as well as monthly average precipitation for January, April, June, July and November (Fig 4.2).

In terms of future projections, fall precipitation continues to show the strongest and most consistent increase across all time periods and scenarios, by up to 30% by end-of-century under higher emissions averaged across all five weather stations. This suggests that the climate models may be accurately capturing the regional factors responsible for observed increases in fall precipitation over the last 50 years. There is also some indication that precipitation may increase in winter, particularly over the near term and under lower emissions. Projections for the spring and summer, however, have large error bars indicating lack of inter-model agreement regarding the magnitude and even the sign of the change (Fig. 4.3).

Seasonal changes in precipitation in the Mobile Bay region are consistent with those projected to occur across the southeastern U.S. and the Gulf Coast (Figure 4.4). Precipitation changes are largest and most consistent during the fall season; winter shows a slight increase under lower amounts of change; changes in spring are inconsistent; and summer shows risk of drying that increases over time and with larger global change. Uncertainties in these projections are discussed further in Section 5.

Similar to figure 3.1, this figure shows the change in seasonal precipitation in the Southeast under global mean temperature increases of 1oC (top), 2oC (middle) and 3oC (bottom), as simulated by the ten climate models used in this analysis.  The level of change varies across the region:  for 1 C change, it ranges from -10% to +10%.  For 2 C, it ranges from about -20% to +15%.  For 3 C, it ranges from about -20% to +30%.

Figure 4.4 Change in Southeast seasonal average precipitation, in percentage relative to 1990-2009, as projected under global mean temperature increases of 1oC (top), 2oC (middle) and 3oC (bottom). Values shown here are the means as simulated by the ten climate models used in this analysis. Projected changes for the Mobile Bay region are consistent with the broader changes projected to occur across the Gulf Coast.

Heavy Precipitation Events

Heavy rainfall events can damage homes, businesses, and public infrastructure. The frequency of occurrence of more than 2 inches of rain in 24h, for example, has already increased across much of the U.S., particularly in the Northeast and Midwest (USGCRP, 2009).

In the Mobile Bay region, exceedence thresholds for 24h annual maximum daily precipitation, as well as cumulative 48h and 96h precipitation, show the most significant and consistent changes from 1960 to 2010. For the maximum 24h precipitation, increases in all the exceedence probabilities from 0.2 to 50% are significant and consistent across all 5 weather stations. For 48h precipitation, increases up to the 5% exceedence are consistent across all stations, and increases in exceedences up to 50% are consistent across four out of the 5 stations. Similar increases are also seen in 96h precipitation (Fig. 4.5).

This figure assesses the trend in observed precipitation from 1960-2010 for different exceedence threshold events in Mobile.  The magnitude of the precipitation amount for many of the events has increased.  For the maximum 24h precipitation, increases in all the exceedence probabilities from 0.2 to 50% are significant and consistent across all 5 weather stations. For 48h precipitation, increases up to the 5% exceedence are consistent across all stations, and increases in exceedences up to 50% are consistent across four out of the 5 stations. Similar increases are also seen in 96h precipitation.

Figure 4.5 Significant increases in many exceedence thresholds for cumulative 24h, 48h, and 96h precipitation are observed at all five stations (dark blue) and at four out of five stations (light blue) over the period 1960 to 2010. No negative values, or decreases, were detected.

In many regions, the observed trend in heavy rainfall is expected to continue in the future as warming temperatures accelerate the hydrological cycle at both the local and global scale (e.g. Tebaldi et al., 2006). Here, projected changes in heavy rainfall events are captured by calculating the exceedence thresholds for the 24h annual maximum and cumulative 2-day and 4-day precipitation from the 0.2nd to the 50th quantile, and the maximum 3-day precipitation totals for each season.

This section focuses on projected changes in the 1st and 50th exceedence thresholds for annual maximum 24h precipitation and cumulative 96h precipitation, as representative of the range of changes generally projected for all precipitation indicators. These two metrics not only average precipitation over a different number of hours, but are also defined slightly differently. Exceedance thresholds discussed here are taken from four distinct 30-year datasets and should not be compared with recurrence intervals usually applied to much larger observed historical datasets. Details regarding the calculations methods used for each metric are provided in the Appendix. Additional projections for the complete set of indicators are available in an Appendix to the Task 2 report.

Most precipitation exceedence thresholds are projected to increase, consistent with observed historical trends. There is some indication of slightly greater increases for shorter duration events, but there is also greater uncertainty in the statistics for these shorter events (Figure 4.6). In general, however, most projections show little difference across scenario, rainfall event duration (24, 48, or 96h), or time period (near-term, mid-century, and end-of-century).

Observed cumulative 24h precipitation for the 5 Mobile Bay weather stations averages around 13.5" for the first percentile and 5" for the 50th percentile. Historical simulated values are slightly lower, averaging 12.5" and 4.5" (a further discussion of biases in simulated historical values is provided in Section 5). Consistent with historical trends, increases in the 1st percentile are projected to continue across all scenarios, all models, and for all future time periods (Fig. 4.6a). There is no real difference between the changes projected to occur under any given scenario, or even for any given time period. Thus, rather than showing as a linear trend, these projections suggest more of a "step"-type increase in the exceedence threshold of the 1st percentile. Similarly, exceedence thresholds for the 50th percentile of 24h precipitation are also projected to increase, although changes are less significant compared to historical values. Again, there is little difference between the changes projected under any given scenario or even for any given time period (Fig. 4.6b).

Figure 4.6a - Click for details(a)Figure 4.6b - Click for details(b)

Figure 4.6c - Click for details(c)Figure 4.6d - Click for details(d)

Figure 4.6 Precipitation exceedence thresholds across most time periods including 24h (a, b) and 96h (c, d) are projected to increase. Increases tend to be slightly larger for lower exceedence thresholds as compared to higher ones (here, 1% compared to 50%), but in general there is little difference between the magnitude of changes projected under different scenarios and for different future time periods. For each time period, the scenarios shown (from left to right) are B1, A2, and A1Fi.

This figure shows maximum three-day precipitation totals for a historic perios (1980-2009), which averages around 5 inches, and three future periods, 2010-2039, 2040-2069, and 2070-2099.  Projected changes averaged across models for each scenario range between 5.5 and 6.0 inches.  As is the case for the other measures of accumulated precipitation shown in Figure 4.6, projections indicate a likely increase in average values of maximum three-day accumulated precipitation, with little distinction between the changes projected under different scenarios or time periods.

Figure 4.7 Maximum three-day precipitation totals currently average around 5 inches per year. As is the case for the other measures of accumulated precipitation shown in Figure 4.6, projections indicate a likely increase in average values of maximum three-day accumulated precipitation, with little distinction between the changes projected under different scenarios or time periods.

Observed cumulative 96h or 4-day precipitation for the 5 Mobile Bay weather stations averages around 7" for the first percentile and 0.65" for the 50th percentile2. Historical simulated values are similar, averaging 8" and 0.6" (a further discussion of biases in simulated historical values is provided in Section 5). Also consistent with historical trends, increases in the 1st percentile are projected to continue across all scenarios, all models, and for all future time periods (Fig. 4.6c). Exceedence thresholds for the 50th percentile of 24h precipitation are also projected to increase slightly, although changes are less significant compared to historical values (Fig 4.6d). As seen for 24h precipitation, there is little difference between the changes projected under any given scenario or even for any given time period, suggesting that the primary driver of uncertainty in projected future changes in precipitation is scientific or model uncertainty (see Section 5 for more discussion on uncertainty).

Average annual maximum three-day precipitation totals were calculated individually for each season. Historical and simulated future annual average values are shown in Figure 4.7. Projected changes are generally positive, although relatively small (averaging less than 1 inch compared to the historical average of 5 inches). Every season shows some indication of an overall increase in the amount of precipitation accumulated during three-day events. However, in most cases the range of projected values includes the potential for no change or even a slight decrease compared to historical simulated values. Seasonal values (not shown, but available in the excel data files accompanying this report) show increases typically on the order of half an inch in spring and summer, and one inch in winter and fall, also with little difference between emission scenarios and time period.


1 The p-value measures the probability of obtaining a given value in error. A p-value of 0.1, therefore, indicates 90% confidence that the trend detected is real. The Kendall tau shows how a given variable is correlated with time in order to demonstrate whether a trend is present. The sign of tau indicates whether the trend is positive or negative (i.e. increasing or decreasing with time). Tau values can range from -1 to +1, with larger absolute values indicating stronger trends.

2 Note that this is a four-day running total, and thus is not directly comparable to the 24 hour data.

Updated: 03/27/2014
HEP Home Planning Environment Real Estate
Federal Highway Administration | 1200 New Jersey Avenue, SE | Washington, DC 20590 | 202-366-4000