Understanding temporal and spatial temperature variation at the local scale in a high latitude environment

  • Gary Pike

    Student thesis: Doctoral Thesis


    The dearth of local scale data in remote high latitude areas means that regional scale data is commonly interpolated to fill the gap. These interpolations have limited accuracy due to the influence of complex topography and resultant decoupling of near-surface temperatures from regional free-air temperatures. Thus relatively little is known about how predicted regional temperature increases over the next century will likely influence local scale climates in subarctic Scandinavia.

    This thesis investigates local temperature variability in the incised valley area where the Kevo Subarctic Research Station (69°45'N, 27°01'E) is located. The main study area covered approximately 20km² encompassing the lake Kevojärvi and three incised valleys and an elevational range of 256m (74-330m a.s.l.). Near-surface temperature data were collected from a network of 60 temperature dataloggers for the period September 2007 to March 2012. NCEP/NCAR reanalysis data were used to reconstruct 6-hourly synoptic conditions for the study area.

    Lapse rates, principal component analysis and regression of surface temperatures on free-air temperatures were used to investigate present local temperature variability. The results were used to infer likely local scale temperature change assuming the strengthening westerlies and storm track predicted for the region. The data from the main network were also used to assess the representativeness of the Kevo Meteorological Station (established 1962). An additional network was set up to collect air and water temperature data from seven nearby lakes in order to validate air temperature estimations for the lakes, and to test a summer lake surface water temperature model based on air temperatures and theoretical solar radiation for remote lakes (June 2010 – September 2011).

    The results show a complex, highly variable temperature structure driven by the high latitude solar geometry, incised topography, variable land cover and synoptic conditions. Inversion conditions dominate temperature variability for most of the year, although with reduced influence during the polar day. Intense and persistent inversion events are a common feature of the winter months with gradients of +80°C/km not uncommon. The strongest inversion gradient was recorded at +92.4°C/km. Clear skies and low winds were the main controls of inversion formation during winter, but due to low temperatures (inhibiting convection) there was a weaker link between anticyclonic conditions and clear skies. Albedo induced steepening of lapse rates commonly gave day time lapse rates beyond -9.8°C/km (dry adiabatic) during March-May, with a peak of - 17.2°C/km measured. Diurnal temperature ranges for south facing low elevation sites were particularly large during the spring due to the daily inversion formation/destruction cycle (up to 30°C). During the ice-free period Kevojärvi had a significant influence on temperatures at low and mid elevations within the study area, resulting in a reversal of the usual day time strengthening of lapse rates. Due to its location in the valley bottom adjacent to Kevojärvi the meteorological station was not located in the most representative place in the local area (ranked 14/61), and certainly not representative of the wider area. The lake surface water temperature model showed good potential for future application. It was thought likely that predicted future synoptic changes in the region would act to inhibit inversion formation and intensity and so result in increased warming in the valley bottoms compared to the hilltops. Finally, due to the high frequency of inversion conditions and the resultant skew in temperature distributions, the reporting of lapse rates for areas such as this should be reported using median values, as mean values can be markedly different and therefore misleading.
    Date of AwardJul 2013
    Original languageEnglish
    SupervisorNick Pepin (Supervisor), Graham P. Wilson (Supervisor) & Philip Soar (Supervisor)

    Cite this