Monitoring a Small Town's Urban Heat Island by Augmenting the Kentucky Mesonet

This study describes how urbanization on Eastern Kentucky University's Richmond campus in Madison County, Kentucky affects air temperature and humidity within the urban canopy layer. This effort is made possible by a partnership between the Kentucky Mesonet and Eastern Kentucky University (EKU). The Kentucky Mesonet is a network of automated weather and climate monitoring stations being developed in the Commonwealth of Kentucky (kymesonet.org). A mesonet station was established at Meadowbrook Farm (MBF) in rural Madison County in 2009. The site is located in a non-irrigated pasture approximately 13km east of downtown Richmond. In 2010, external funding was obtained to install a “sister station” in the heart of EKU's Richmond campus. Simultaneous air temperature and humidity observations from the two stations were compared to assess the urban heat island effect in the urban canopy layer (DT=Turban-Trural); the station at MBF being ‘rural' and the station at EKU being ‘urban'. Observations collected at each site from August 2010 through February 2013 were used in this analysis. Each day's max/min temperature was given by the highest/lowest 5-minute temperature observation at each station that day. Monthly averages of daily maximum temperature at the two stations were within ±0.5°C of each other in 28 of 31 months. Monthly averages of daily minimum temperature, however, were at least 1°C higher at EKU than at MBF in 30 of 31 months. The largest difference in monthly average daily minimum temperature was 4.2°C, which occurred in November 2012. Hourly values of UHI intensity and heating/cooling rates were calculated after averaging 5-minute temperature observations into hourly values. This analysis reveals an urban heat island that has a diurnal pattern similar to many other UHIs reported in the literature. The diurnal cycle of the UHI is explained by differences in heating/cooling rates at the two station locations. Each afternoon, the rural area begins cooling before the urban area, and achieves a greater rate of cooling than the urban area. Thus, an urban heat island begins developing just before sunset and continues to intensify up to a few hours after sunset. At that point, cooling rates at both sites slow substantially. Cooling at the rural site ultimately slows so much that cooling at the urban site exceeds it, causing a decline in UHI intensity through the rest of the night. At sunrise, both sites warm rapidly. Greater heating rates occurring in the rural area causes a sharp decline in UHI intensity after sunrise. UHI intensity remains relatively small until the late afternoon when the process repeats itself. The largest positive hourly UHI intensity (EKU warmer than MBF) was 13.24°C on 11/12/10 at 10pmEST. The largest negative UHI intensity was -3.31°C on 2/27/12 at 5amEST. Hourly temperatures at the two stations were within ±0.5°C of each other 62.9% of the time during the day and during 31.4% of nocturnal hours. Positive UHI intensities >0.5°C occurred in 27.2% of daytime hours and 58.9% of nighttime hours. Weather has a pronounced effect on EKU's UHI as the most intense heat islands occur during anticyclonic conditions with light winds. Seasonal differences in synoptic conditions are at least partly responsible for seasonal differences in UHI behavior. Intense UHIs may develop at any time of the year, but average UHI intensity is greater during the warm part of the year when synoptic conditions favoring UHI development are more common. From August 2010 through February 2013, average nocturnal UHI intensity was 0.81°C in winter, 1.62°C in spring, 2.65°C in summer, and 2.63°C in fall. The urban location was found to be less humid than the rural location on average as dew point was almost always higher at MBF than at EKU. Dew point was more than 0.5°C higher at the farm than on campus in 88.6% of all daytime hours and 74.7% percent of nighttime hours. Compared to MBF, EKU is relatively drier during the day than during the night. This is especially true during the warmest part of the year when daytime differences in dew point are largest (i.e., EKU is driest relative to MBF on summer days). Seasonally, average daytime dew point differences between the two stations were 1.15°C in winter, 1.69°C in spring, 1.85°C in summer, and 1.72°C in fall. The magnitude of the differences between simultaneous temperature and humidity observations at the two stations is somewhat surprising given Richmond's relatively small population (31,364 in 2010 [census.org]) and abundant green space on campus. Positive UHI magnitudes (TEKU more than 0.5°C warmer than TMBF) are most common during the warm season and average nocturnal UHI intensity is greatest in summer as well. Heat islands during this part of the year increase summer cooling costs and degrade outdoor comfort. These impacts should be assessed and potential mitigation strategies investigated. The substantial difference in daily minimum temperatures between the two stations has implications for long-term climate records and studying climate change. This case is particularly interesting since the location of the campus site was the location of a NWS COOP station from 1893-2001. The old cotton region shelter still stands just a few meters from the new tower. Future work should examine the old station's data record for urban bias. Establishing the automated weather station at EKU has attained at least one unforeseen benefit. Numerous urban flooding events have occurred in downtown Richmond during the past few years. CDP Engineers of Lexington, KY has been hired to address this issue and 5-minute precipitation data from the EKU station has been a key piece in their efforts. The productive partnership between the Kentucky Mesonet and EKU takes advantage of the fact that the Kentucky Mesonet is a scalable network. Instrumentation identical to all mesonet stations was purchased for the EKU station so that its data can be directly comparable to other stations in the network. Moreover, using the same equipment allows data from the EKU station to be seamlessly gathered and archived by the mesonet office. Similar opportunities to expand existing observational networks through cooperative efforts exist and should be sought out for specific environmental monitoring needs.