Analysis of nocturnal air temperature in districts using mobile measurements and a cooling indicator

The urban heat island phenomenon is generally defined as an air temperature difference between a city center and the non-urbanized rural areas nearby. However, this description does not encompass the intra-urban temperature differences that exist between neighborhoods in a city. This study investigates the air temperature dynamics of neighborhoods for meteorological conditions that lead to important urban heat island amplitude. Local climate zones (LCZs) have been determined in Nancy, France, and mobile screen-height air temperature measurements are performed using an instrumented vehicle. Initially, hourly measurements are performed within four different LCZs. These results show that air temperature within LCZ demonstrates a nocturnal cooling in two phases, i.e., a first phase between 1 to 3 h before sunset and 3 to 5 h after sunset, and a second phase from 3 to 5 h after sunset to sunrise. During phase 1, neighborhoods exhibit different cooling rate values and air temperature gaps develop between districts, while during phase 2, cooling rates tend to be analogous. Then, a larger meteorological data set is used to investigate these two phases for a selection of 13 LCZs. Normalized cooling rates are calculated between daytime measures and nighttime measures in order to quantify the air temperature dynamics of the studied areas during phase 1. Considering this indicator, three groups are emerging: LCZ compact midrise and open midrise with mean normalized cooling rate values of 0.09 h (-1) LCZ large lowrise and open lowrise/sparsely built with mean normalized cooling rate values of 0.011 h (-1) LCZ low plants with mean normalized cooling rate values of 0.014 h (-1) Results indicate that the relative position of LCZ within the conurbation does not drive air temperature dynamics during phase 1. In addition, measures performed during phase 2 tend to illustrate that cooling rates are similar to all LCZ during this period.