Nick Bond, the Washington State Climatologist, pointed out an interesting observation at a meeting I attended last week. For summers from 2000-2010, nighttime temperatures (T-min) in many locations in the Pacific Northwest have shown a strong warming trend while daytime temperatures (T-max) have shown a general cooling trend (Panel 1).   Each circle on the map is scaled based on the station’s temperature trend with red indicating increasing temperature and blue decreasing. This could be part of the reason I’ve had trouble sleeping at night in recent summers – more below!

Panel 1: Summer Min (nighttime) and Max (daytime) Temperatures: 2000-2010;   Source:

The long-term trend (1910-2010) is similar, for both annual (Panel 2) and summer temperatures (Panel 3) with nighttime temperatures trending warmer and daytime temperatures trending cooler or trending warmer to a lesser degree than nighttime. The intermediate-term trend (1960-2010) also follows the general pattern for summer temperatures, though the contrast is less stark (Panel 4).


Panel 2Panel 2: Annual Min (nighttime) and Max (daytime) Temperatures: 1910-2010;   Source:

Panel 3Panel 3: Summer Min (nighttime) and Max (daytime) Temperatures: 1910-2010;   Source:

Panel 4Panel 4: Summer Min (nighttime) and Max (daytime) Temperatures: 1960-2010;   Source:

I would encourage you to go to the State Climatologist Office website and play with this tool a bit after reading the tutorial. Not only can you change the parameters as I have presented in these panels, but you can select a specific weather station and graph the data specific to that station. I selected three time periods to present (short, intermediate and long) focusing primarily on the summertime temperature differentials I was interested in. If you select different time periods (or seasons, or months), you’ll notice that the pattern I’ve showed above doesn’t always hold true (for instance 1980-2010 summertime doesn’t show much contrast between changes in T-max or T-min).

This raises an important point. We have to be extremely careful when interpreting a trend and not over-state the data or our expectations of what it means for the future. From the panels above, we can interpret that for the short (decadal), intermediate (half-century) and long-term (century) there has been a trend toward greater summertime increases in nighttime temperatures (T-min) than daytime temperatures (T-max) in the region, but the magnitude of change (increase in temperature) and contrast (difference between day and night) are not linear for the selected time periods. There is also a lot of noise in the data on a year to year basis. For instance, just looking at the Ellensburg station, there has been about a 0.43 degree F increase per decade in summer T-min over the century, but nearly double that increase (0.8 degree F) just from 2000-2010. During the same decade, the Cle Elum station “next door” experienced about the same magnitude decrease in summer T-min. Furthermore, any single year of the last decade may have been above or below the trend. So, depending on where you are, nighttime temperatures may or may not be a good explanation for summer insomnia.

When considering what this trend might mean for agriculture, there are two immediate thoughts that come to mind. First, for livestock production, particularly large mammals like cattle, the effects of this trend are likely to be similar in concept to what we humans experience: more discomfort and stress during warm summer nights. Mauger et al. (in press) found that milk production is likely to be reduced by increased temperatures, particularly during summer in regions with more severe summer temperature and humidity impacts like Florida. One reason Washington is a relatively good place for dairy farming is that our low-humidity climate during summer tends to provide a lot of relief to the animal (or human) even when daytime temperatures approach 100 degrees F.

We also know that increased nighttime temperatures during summer can affect crops. A 2004 experiment conducted at the International Rice Research Institute and published in the Proceedings of the National Academy of Sciences (Peng 2004) found that rice yields were decreased by 10% for each 1 degree C increase in nighttime temperature during the growing season while increases in daytime temperature did not have an effect.

Our efforts to date to project crop yields under future climate do include consideration for projected changes in T-max and T-min, but we have only presented and evaluated the results for aggregate changes in temperature. We have yet to really tease out the relative effect of changes in T-min vs. T-max, but based on the observations above this is something that we’ll start to look at more closely.



Mauger, GS, Y Bauman, TD Nennich, and EP Salathé: Impacts of Climate Change on Milk Production in the United States. American Geographer (in press).

Peng S, Huang J., Sheehy JE, Laza RC, Visperas RM, Zhong X, Centeno GS, Khush GS, Cassman KG. (2004). Rice yields decline with higher night temperature from global warming. PNAS. 2004 101(27):9971-5.

Washington State Climate Office N.W. Temperature, Precipitation, & SWE Trend Analysis tool.