Wednesday, August 6, 2014

Barrow-Area Temperatures

The Barrow Climate Reference Network (CRN) monitoring site has been in operation for 12 years now, and I thought it would be interesting to look at the observed temperature differences between the CRN location and the airport ASOS location.  The CRN station is located about 4.5 miles northeast of the airport station, closer to Point Barrow; the red stars mark the locations on the map below.

To begin with, we know that there is a pronounced urban heat island effect in winter in Barrow (see the 2003 study here), so we expect to see warmer winter temperatures at the airport location.  The 2002-present CRN data confirm this, but also show that the temperature difference between these two locations persists year-round on average.  The charts below compare the monthly means of the daily maximum and minimum temperatures:

The maximum and minimum temperature differences show substantial seasonal variations; June and July show warmer maximum temperatures at the airport (more than 2 °F warmer on average), but minimum temperature differences are largest in winter, especially late winter.  Interestingly, the Hinkel et al study linked above reported cooler temperatures in the urban area in summer compared to their rural measurements, but their rural zone was well to the southeast where the maritime influence was less.  Both the Barrow CRN and Barrow ASOS are close to the coast and so the urban heat island effect remains evident in summer.

The frequency distribution of temperature differences is illustrated in the following two charts, which show that daily maximum temperatures are lower at the airport less than 10 percent of the time year-round, while daily minimum temperatures are lower about 15 percent of the time.  Note that the daily maximum and minimum temperatures are defined for the same midnight-to-midnight period (standard time) at both stations.

The propensity for much higher summer maximum temperatures at the airport, and much higher winter minimum temperatures, is also clearly seen in the charts.  For example, the airport is 5 °F or more warmer by day on 16 percent of summer days, while minimum temperatures are 5 °F or more warmer in Dec-Feb on 10 percent of days.  I haven't looked at the dependence on wind speed, but it's likely that the temperature differences are much larger on relatively calm days, regardless of season (as found by Hinkel et al).

Finally, it's interesting to look at the annual mean values over the 11 complete years of overlapping data - see below.  The blue bars show the airport annual mean maximum and minimum temperatures, and the red bars show the CRN data.  Over this short period there is an upward trend in the annual mean temperature difference between the stations, and further investigation of this might be warranted.  An increase in the urban heat island effect seems a bit unlikely, as the population of Barrow dropped by 10 percent between 2000 and 2010 according to the U.S. census data.  It's possible that the weather pattern has favored less windy conditions in the most recent years, or perhaps the change is related to a minor relocation of the airport observing equipment in April 2008 (according to NCDC metadata records).  This latter possibility illustrates a major part of the rationale for the CRN program - with the CRN data, we don't have to worry about local changes to the environment, equipment upgrades, or faulty sensors.  It is a wonderful source of climate data and will only get better as more stations are added and as the years go by.


  1. I wonder if the Chinook-like winds into Barrow that Gary has mentioned before have an effect. And it would be interesting to see how much the temp differences are influenced by solar radiation/cloud cover. Ice breakup is different for different areas. How does this effect the stations? It just seems to me that there are physical land effects that are more influential than we think.

    1. There is certainly a lot more useful work that could be done to examine the physical phenomena that are at play. Cloud cover would be an interesting one for a future blog post. As for wind speed, Hinkel showed a very strong dependence of the winter UHI magnitude on wind speed (their Figure 7); strong winds from any direction will tend to mix out the boundary layer and reduce spatial temperature gradients.

  2. Wind direction and intensity probably effect the observed temperatures. If the fetch is over warm dark land, then one temperature scenario prevails. If over a relatively cooler sea or reflective ice, then another.

    If flow stagnation occurs, then the effect of a local heat island (dark runway and taxiways?), cloud cover, and insolation may determine the temperatures. At least that's been my experience along coastal areas in Alaska.

    Forget Barrow for the moment and consider sea breezes in general:


  3. By looking at the map I would think that the Airport would be slightly warmer with a west wind due to more flow over land (warmer??), while an east wind would bring fairly similar temperatures due to the proximity to the ocean. MKK

    1. MKK, I would say that winds from either due west or due east will both bring a maritime influence, but a southerly flow would be continental in nature. I'll see if I can look at the temperature differences by both wind speed and direction; there should be some interesting variations.