Friday, July 26, 2013

Intricacies of Local Temperature Trends, Part II

In the previous post (Intricacies of Local Temperature Trends, Part I), we showed that average summertime temperatures have increased in recent decades (since 1948), but also that minimum temperatures are increasing faster than maximum temperatures, resulting in a decrease in what meteorologists call the diurnal (daily) temperature range (DTR).

There are potentially a number of mechanisms contributing to these trends.  These mechanisms include:

1. Global warming.  It is well documented that average global surface temperatures have increased in recent decades and that this has been accompanied by a decrease in the DTR at many stations.

During this period, the planet's energy balance has been modified by changes in greenhouse gas concentrations, aerosol (i.e., pollution) concentrations, and other factors.  Increasing temperatures and a decreasing DTR are consistent with what we expect from the net influence of these changes, but trends at any given location such as the Salt Lake City airport are also going to be influenced by regional climate variability and related changes in the large-scale atmospheric circulation, cloud cover, and aerosol concentrations.

Thus, we can conclude that trends at the Salt Lake City airport are consistent with expectations from global warming, but the size of the contribution remains undetermined.  I suspect it is a significant, but that other factors noted below are also important.

2. Urbanization. The urban heat island typically acts to increase temperatures and reduce the DTR in developed areas compared to rural or undeveloped areas.  The urban heat island of the Salt Lake Valley is somewhat unique, however, since in addition to adding buildings and the like, we have replaced the natural, semi-arid landscape with an urban forest (i.e., grass, trees, etc.) that is heavily irrigated.

Urban tree canopy in 2006.  Source: Lowry (2010)
Thus, although many urban areas are characterized by less evaporation and transpiration compared to the surrounding rural or undeveloped region, evaporation and transpiration are greater in the Salt Lake Valley than they would be with a natural landscape, resulting in a cooling oasis effect.

So, we have traditional heat island effects that would contribute to higher temperatures (maximum and minimum) and a reduced DTR compared to an undeveloped Salt Lake Valley, but this is countered somewhat by the oasis effect.  In the net, I'm not sure what wins for the maximum temperature, but it is likely that urbanization in the Salt Lake Valley would contribute to higher minimum temperatures.

The urban area of the Salt Lake Valley has grown considerably over the past few decades, with the population of Salt Lake County increasing from 462,000 in 1970 to 1,049,000 in 2011.  This has been accompanied by considerable expansion of development into the southern and western Salt Lake Valley.

Source: Lowry (2010)
The Salt Lake City Airport lies at the edge of the urban area, so it probably isn't located where the urban heat island is the strongest.  That being said, traditionally at the time of minimum temperatures, the low level flow is southeasterly and from the urban area to the airport.  Thus, some of the increase in minimum temperature and decrease in DTR over the past few decades is probably related to urbanization.  The influence on maximum temperature (in the summer) is less clear. 

3. Airport and instrumentation changes.  Changes in the local characteristics of an observing site can be extremely important, as discussed in this recent post on Cliff Mass's blog.  Moving a site even a short distance, altering the landscape or development around it, or changing the instrumentation can result in major changes in the temperature statistics.  The Salt Lake City temperature time series doesn't seen to show any major jumps (up or down) in temperature, but that doesn't mean there isn't a fly embedded in the ointment.  I've never dug into the records enough to address this concern, but it is an issue worthy of investigation. 

4. Wildcards.   There are other factors too numerous to discuss, and all of these factors can interact in unusual ways.  For instance, a small increase in incoming long-wave radiation at night can sometimes have a big impact on minimum temperature if it weakens the stability of the boundary layer just enough to allow the turbulence to mix the surface layer.  Sometimes a small change in the surface energy balance has a big impact on temperatures, other times not.  Much work is needed to understand these effects.  

In summary, average summertime maximum and minimum temperatures have increased in recent decades at the Salt Lake Airport.  Minimum temperatures have increased more than maximum temperatures, resulting in a decrease in the DTR.  It is likely that these trends reflect global warming and urbanization (with climate variability contributing to variations on shorter time scales), with their relative contribution and possible role of changes in land surface and instrumentation at the airport still unexplored.  I'd say there's a good research topic here, but I also know that teasing out all these interactions is incredibly challenging.   


  1. It might be interesting to do a sensitivity study with a limited area micro-scale model that includes an interactive biosphere parameterization (transpiration from vegetation) that can be turned on and off. You could run an experiment with little vegetation and ground water values as observed west of the city and then another simulation with present day vegetation and present day observed ground water. That would be interesting and might add insight. Of course, you'd need some sort of planetary boundary (PBL) layer parameterization as well where the PBL responds to radiative forcing.

  2. This is a difficult thing to sort out. If you plot other stations in Utah on the xmACIS2 site, many of them rural where you'd perhaps expect less land cover change, they show a lot of variability in min temperature trends...some up, some down, some nearly constant. I wonder how representative the SLC airport is of northern Utah and how good the quality is of other stations in Utah. You'd probably need a lot of high quality stations from around the country to get sample sizes large enough to separate out these effects, but it would be difficult.