SR SSD 2002-10
4/2002

Technical Attachment

1. Introduction

Summer in north Texas, specifically in the Dallas/Fort Worth Metroplex, can be described with one word...HOT!! During the years of 1998, 1999, and 2000, the Metroplex experienced extremely hot and dry summers, with 1998 being the second warmest on record. The summer of 2000 also featured an 84 day rainless streak at DFW International Airport (KDFW), the official observation site for the region.

The summer of 2001 was cooler and wetter than normal, but, as can be expected, there were some heat waves. In early July, the official high temperature for DFW Airport appeared to be cooler than surrounding sites. High temperature data were collected from several Metroplex observation sites in an attempt to determine if there was any actual basis for what seemed obvious to even a casual observer. The fact that DFW Airport's high temperatures were typically cooler than the surrounding locations on many days became the subject of media inquiries and even an article in the local newspaper. It was also obvious that the number of days with temperatures 100 degrees or higher at KDFW were less, and in some case much less, than many other surrounding sites.

2. Methodology

The period of record for this study was July 1-October 20, 2001. For each day, high temperatures were collected from eight locations across the Dallas/Fort Worth Metroplex, which surrounds the DFW International Airport.

The locations chosen were:

Dallas/Fort Worth International Airport (KDFW)
Dallas Love Field (KDAL)
Fort Worth Meacham Airport (KFTW)
Fort Worth Alliance Airport (KAFW)
Arlington Municipal Airport (KGKY)
Denton Municipal Airport (KDTO)
McKinney Municipal Airport (KTKI)
Terrell Municipal Airport (KTRL).

Fort Worth Meacham and Fort Worth Alliance airports are west of KDFW, Arlington Municipal Airport is south, Dallas Love Field and Terrell Airport are east, and Denton and McKinney are north. Also, all of these locations are airports at which the Automated Surface Observing System (ASOS) takes observations 24 hours a day, seven days a week.

In order to eliminate temperature outliers, a comparison was made between the high temperature at KDFW and the average high temperature of the other sites, termed the "seven-site average" (SSA). Across north Texas, the gradient of average annual precipitation varies significantly from west to east, so from the Fort Worth airports (KFTW and KAFW), the two westernmost sites, to Terrell (KTRL), the easternmost location, average annual precipitation varies by almost 20 inches. Western counties in the Fort Worth county warning area receive an average of around 28 inches of rain annually, while a few locations in the east receive 46 inches. Across the Metroplex (described for these purposes as Dallas, Tarrant, Denton, Collin, Rockwall, Kaufman, Johnson, Ellis, Parker and Wise counties) precipitation varies from about 30 inches in the west to 42 inches in the east. Also, there is a sharp drop off in vegetation from east to west over the region as well. In general, summertime high temperatures are warmer to the west, where there is less rain and fewer trees, and cooler in the east, where more rain is received and more vegetation grows. Using the SSA alleviated some of these temperature discontinuities.

There were obviously some variables in this study that were difficult, if not impossible to account for. First, each ASOS unit is sited individually, and some are placed in more representative locations than others. Often placement decisions have to be agreed upon by airport managers and the National Weather Service. Secondly, due to the shear size of Dallas/Fort Worth International Airport, temperatures likely vary from one side of the runway complex to the other. Third, there were several periods of showers and thunderstorms during the study period. If one site received rain while others did not, large temperature differences could be introduced into the dataset.

3. Results

In July, DFW Airport had three high temperatures at or above 100 degrees. Other sites across north Texas had significantly more, ranging from 21 at Alliance Airport, to 18 at Meacham Airport. On the low side, there were four days of 100 plus temperatures at Terrell and five at Arlington. The average number of 100 degree plus temperatures, among these sites in July was ten.

In August the results were similar. DFW Airport only had five days of 100 plus readings, while Alliance had 12 and Meacham ten. Denton, Terrell, and Arlington each had seven. The average number was nine. So, DFW Airport was below average for the number of 100 degree high temperatures in the Metroplex for both months.

What was more pronounced was the comparison of temperatures at DFW Airport to the individual sites (Table 1). In July, KDFW had an average high temperature of 96.7 degrees. Denton had an average of 97.7 degrees, Alliance 99.5, Meacham, 98.7, Arlington 96.8, Dallas Love Field 97.8, McKinney 97.2 and Terrell 97.1. The difference between DFW's average and the individual sites ranged from -2.8 degrees at Alliance Airport, -2.0 at Meacham, to -0.1 at Arlington. Comparing the DFW average (96.7 degrees) to the seven-site average (SSA), DFW was still 1.1 degrees cooler.

Table 1

In August, with generally cooler temperatures overall, DFW Airport had an average temperature of 94.1 degrees. Temperatures ranged from 94.3 to 96 for the remaining sites (Table 2). Differences ranged from almost 2 degrees cooler (-1.9) at Alliance, 1.1 degrees cooler than Meacham to barely a quarter degree (-0.2) cooler than Arlington.

Table 2

With the hot temperatures ending by mid-August, another interesting question was if there were any differences among the sites in the autumn season. In September, there were no 100 degree plus temperatures at any of the studied sites. DFW had an average of 83.7. For the first time in this study, DFW's average high temperature was actually warmer than a few individual sites. However, it still remained cooler than the SSA. Most of the individual temperatures were close, and all were within plus or minus one degree. Denton had an average of 83.7 degrees, Alliance 83.7, Meacham 83.4, Arlington 84, Dallas Love Field 84.3, McKinney 82.9 and Terrell 83.4. (Table 3) The differences ranged from DFW cooler than KDAL by 0.6 degrees, but warmer than McKinney (0.8 degrees) and Terrell (0.3 degrees).

Table 3

Results in October were similar to those in September. The average high temperature at DFW was 78.0 degrees. Averages across the region ranged from 77 degrees at McKinney to 78.5 at Denton. Differences ranged from DFW cooler than Denton (-0.5 degrees), Dallas Love Field (-0.39 degrees), Alliance and Arlington (-0.2 degrees) and Meacham (-0.4 degrees) and warmer than McKinney (+1.0 degrees) and Terrell (+0.2 degrees). (Table 4)

Table 4

One interesting conclusion was that during July, when the SSA was at or above 100 degrees, DFW was cooler than the area average by 1.7 degrees. As the SSA cooled, the cool bias at DFW diminished. For example, when the SSA was 98, the difference was 1.5 degrees cooler, 96 brought a difference of 1.2, 94 (-1.20) and 90 (-1.0).

In August the difference was -1.3 for an SSA at or above 100 degrees; -1.2 at 98 degrees; -1.0 at 96 degrees; -0.90 at 94 degrees. From 94 degrees downward to 78 degrees the variances were less than a degree. While DFW was always cooler than the average, the differences even increased slightly in a few temperatures, if rain or other factors created an outlier which slightly skewed the average.

For the month of September, there was much less variability between DFW's high temperature and the SSA. However, the differences were greater for higher temperatures. With an average high temperature greater than 90 degrees, the difference was actually shown to have DFW warmer than the SSA by 0.9 degrees in September. DFW's difference was warmer than the average SSA difference for highs ranging from 90 to 70 degrees. At 80 degrees, the difference was only 0.04 degrees. Once again rain and clouds the first part of the month led to some days with temperatures at one site being significantly cooler and creating outliers in the data.

October showed small variations, with DFW almost one-half degree cooler (-0.4) than the SSA at or above 86 degrees and one degree cooler at 84 degrees. From 82 to 68 degrees the variations had DFW cooler but by 0.2 degree or less. At 66 and 64 degrees, the monthly difference at DFW was actually warmer than the SSA average variation but again by less than 0.2 degree.

4. Discussion

It is quite evident from this study that during the summer of 2001 the ASOS temperature sensor at Dallas/Fort Worth International Airport read cooler in comparison to surrounding ASOS locations. This cool bias was generally greater with increasing temperatures. While the differences between DFW's high temperature and individual stations were often larger than the difference between DFW Airport and the seven-site average (SSA), the same conclusions can be drawn. DFW was almost two (1.7) degrees cooler than the SSA in July, while variances between DFW and Fort Worth Alliance Airport were almost three degrees (DFW 2.8 degrees cooler). July was the hottest month of those studied. With temperatures at or above 100 degrees, DFW was 1.7 degrees cooler than the SSA. When using temperatures at or above 96 degrees, the difference dropped to DFW being 1.2 degrees cooler than the SSA. As temperatures continued to decrease in late August through October, the variances were much less, decreasing to around 0.1 degree when highs were in the 60s and 70s. There were times due to localized rain or cloud cover at surrounding locations which allowed DFW's high temperature to be greater than the SSA.

National Weather Service regulations allow for a plus or minus two degree variation in temperature between the automated ASOS sensor and a reference manual observation. None of the lower readings at DFW Airport were out of this tolerance. However, it does seem that there is valid evidence to draw the conclusion that the recorded temperature at DFW International Airport is generally cooler than the surrounding locations taken separately or in average. This variance is more pronounced when temperatures are warmer, with the larger cool biases occurring when temperatures exceed 100 degrees.