Margaret LeMone | 21 September 2009 • When our family bought our last car, we debated about the color, wondering if a white car would be cooler on our summer road trips across the American West than a darker-colored car. The family was divided on whether the color would matter, but we ended up purchasing the white car. But was it for the right reason? I was reminded of this debate recently when I heard that the state of California was thinking of passing laws essentially requiring lighter-color cars, the argument being that the dark cars would be heated more, requiring more air conditioning and thus wasting energy unnecessarily.
My colleagues at NCAR were divided into two camps: "dark is hotter" and "color makes no difference." A person in the latter camp based his impression on an informal experiment for which he installed thermometers in a dark-colored car and a light-colored car and saw no significant difference.
Thus three of us began an experiment to compare the temperature at various times of day inside two cars that were identical except for the paint color. The first step was to find two such cars. The two Toyota 4Runners, pictured below, met these criteria pretty well. (Note that the apparent difference in the shading of the windows is due to the photographic exposure rather than reality, as can be seen from comparing the color in the two backgrounds.)
Note to readers: This experiment was done in the spring, before the temperatures outside got very warm. During the summertime, temperatures inside parked cars can get much warmer - and are dangerous for anyone inside, especially pets and small children.
In addition to being the same type of car, the two cars both had gray upholstery (though one was cloth and the other leather). The front and back windows were tinted on both cars. All the windows on the white car were tinted, while the front side windows on the black car were not tinted.
We measured the temperature in the second row of seats, with the thermometers placed such that they were shaded from the sun. Also, we made sure that the thermometers were in the same location for both cars. Thus, for example, both thermometers would be placed on the floor behind the right front seat. The thermometers were left in the cars for the duration of each comparison, unless otherwise noted. When we measured the temperatures, we confirmed the thermometer was in the shade, observed cloud cover and type (since clouds affect sunlight hitting the cars) and recorded the wind speed (since we thought wind could keep the outside of the car cooler, and thus the inside cooler as well). There was more than one observer. Some would estimate the wind speed using the Beaufort scale, while others took the wind speed measured at the Foothills Lab weather station located on the roof of the building next to the lot where the cars were parked.
We started with some thermometers I found around the house, a "square" thermometer that came from our refrigerator (only goes up to 90°F), and a "round" thermometer that we used to measure the temperature inside the house (goes up to 120°F). In general, these thermometers agreed with one another quite well at temperatures around 80°F and below, but they departed at warmer temperatures (Table 1). Furthermore, the square one seemed to adjust more quickly, which meant that the differences between the two were not always in a consistent direction.
Table 1. Comparison temperatures (degrees F) for thermometers I brought from home. The square thermometer is a refrigerator thermometer with a scale to 90°F; the round thermometer has a scale that goes up to 120°F.
We started our car-comparison experiment by pointing both of the Toyota 4Runners toward the west. In this case, the black car was slightly warmer. However, this could be due to the untinted front side windows in the black car as well as the paint color (see Table 2).
Table 2. Temperatures (°F) inside cars, pointed west. Measurements taken during morning.
When we pointed the cars toward the south, the black car was cooler more often than the white car, but we couldn’t find a strong, consistent difference using the square and circular thermometers I brought from home–at least to within the accuracy of the thermometers. The differences in time response led to biases that were not consistent, so it is hard to justify “correcting” the data. We believe that this lack of consistent difference is due to the majority of the sunlight coming in through the identical front windows.
Table 3. Temperature (°F) inside cars, pointed south (Sq = square, O = circular)
Because of this bias, we replaced the round and square thermometers with two alcohol thermometers obtained from NCAR’s Earth Observing Laboratory (EOL), obtaining the data in Table 4. Here, we saw a significant difference between the cars during the morning, when it was sunnier. As for the first entry on Table 3, part of the difference could be due to window tint in the white car–at 10:50 MDT (9:50 MST), a significant amount of the sunlight passes through the side windows. In the afternoon, not only was the Sun closer to zenith, but there were more clouds. Note that the white thermometer was biased high by 0.5°F. The ‘white’ temperatures were not corrected for this.
Table 4. Temperature (°F) inside cars, pointed south, with NCAR/EOL thermometers
When the cars were pointed north (Table 5), the black car was definitely warmer, especially in the afternoon when skies were clear. Again, the away-from-solar-zenith temperature differences are probably related to the differences in tinting as well as the color of the car, but for the observations near solar noon (1310 MDT, 1330 MDT) the differences are probably related to the paint color.
Table 5. Temperature (°F) inside cars pointed north, with NCAR/EOL thermometers. Comparisons to square (Sq) and round (O) thermometers included.
I am almost ready to conclude that our family argument went the right way, and maybe the state of California is right, but we want to do one final test: drive both cars and see how the temperature changes if the air conditioner is not turned on. After all, people use energy when they are driving–not when the cars are parked. I’ll cover the results of this test in a future post.