Identification of mid-latitudinal regional and urban temperature variabilities based on regional reanalysis data

Abstract

The objective of the study is to detect geographical and temporal variations of near surface air temperatures over Minnesota and Wisconsin, USA derived from the North American Regional Reanalysis (NARR) dataset. In addition, the study serves to assess the usefulness of NARR temperature data to analyze regional and local temperature variations. Particular emphasis was placed on the analyses on the temperature-modifying effects of the Great Lakes and large urban environments. We analyzed annual mean, daily maximum and minimum, and January minimum and July maximum temperatures for the period 1979–2006 by using methods such as ordinary kriging, principal component analysis, and the Mann–Kendall test. On a regional scale, we found significant effects of the latitude and the Great Lakes on the spatial variability of the data. Furthermore, we found clearly identifiable effects of large urban areas in the study region (Minneapolis—Saint Paul and Milwaukee), which are more evident in the principal component scores than in the temperature data themselves. While we failed to detect significant July maximum temperature trends, we detected significantly increasing trends in January minimum and mean annual temperature datasets in the eastern part of the region. Overall, the present study has demonstrated the potential of using NARR data for urban climate research.