Date of Award

1-1-2022

Language

English

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

College/School/Department

Department of Atmospheric and Environmental Sciences

Content Description

1 online resource (vii, 172 pages) : color illustrations, color maps.

Dissertation/Thesis Chair

Aiguo Dai

Committee Members

Brian EJ Rose, Paul E Roundy, Liming Zhou

Keywords

Bengaluru, India, MODIS, Urban heat island, Urbanization, WRF, Humidity, Heat waves (Meteorology), Atmospheric temperature, Vegetation and climate

Subject Categories

Atmospheric Sciences | Environmental Sciences

Abstract

The urban heat island (UHI) effect refers to how urban surfaces tend to be warmer than nearby non-urban areas due to less vegetation and other processes. UHIs can increase the risk of heat and respiratory illnesses. Since every city is unique, UHIs should be studied on a local-scale. One particular city that has not had its UHI comprehensively evaluated is Bengaluru, India. Bengaluru was once known as the “Garden City” of India due to a wide presence of gardens and public parks, but is now known as the “Silicon City” of India due to the overwhelming presence of the information technology industry. This dissertation aims to investigate Bengaluru’s UHI. First, the UHI was analyzed during the dry (December-January-February; DJF) and wet (August-September-October; ASO) seasons during day and night using land surface temperature (LST) data from the MODerate Resolution Imaging Spectroradiometer (MODIS). Results showed that the 2003–2018 mean UHI intensity was highest for DJF nighttime (1.43°C), followed by ASO daytime (1.14°C), ASO nighttime (1.02°C), and DJF daytime (–0.60°C). It was hypothesized that increasing urban aerosols may explain the negative UHI in DJF daytime since aerosols can absorb and scatter solar radiation and have a long residence time during the dry season. To better understand the causes of the UHI, an investigation of the relative importance of the leading controlling factors was explored using multiple linear regression and the random forest. The variables analyzed included albedo, aerosol optical depth (AOD), enhanced vegetation index (EVI), latent heat, soil moisture, specific humidity, and wind speed, which were chosen given Bengaluru’s tropical, moisture-rich location and since much of the city used to be covered by vegetation, but now by buildings, and that urban aerosols are increasing. Both approaches showed that EVI is more important than AOD. Therefore, the presence of aerosols is high enough to cancel an UHI that would otherwise occur in DJF daytime due to low vegetation. Next, the Weather Research and Forecasting (WRF) model was evaluated for its ability in simulating LST over Bengaluru and its sensitivity to urban canopy model (UCM) and planetary boundary layer (PBL) schemes. By comparing the simulations to MODIS LST, results showed that urban LST was more sensitive to UCM choice than PBL scheme and the use of an UCM reduced urban LST biases, which led to improved simulations of the UHI. For the best case, urban LST was underestimated by less than 1°C during DJF day and night, and was overestimated by 1.88°C and 0.08°C in ASO day and night. In general, the single-layer UCM (SLUCM) had the least bias for urban LST and UHI intensity. Different UCMs calculate radiative and surface fluxes differently, which could lead to distinct urban LST biases. During daytime, using No-UCM produced a near-zero latent heat flux and the multi-layer UCM (MLUCM) trapped too much shortwave and longwave radiation, both resulting in large, positive urban LST biases. During nighttime, the MLUCM had a negative urban LST bias due to too much longwave radiation reflecting between buildings, causing the lower atmosphere to be warmer than the surface. WRF experiments were then ran with perturbed vegetation cover by changing the control urban fraction of 0.90 by +10%, –10%, –20%, and –30%, where decreases represent greening. The responses in LST, UHI intensity, latent heat (LH), sensible heat (SH), and ground heat (GH) were analyzed. As expected, increased vegetation caused a decrease in LST, UHI intensity, SH, GH, and an increase in LH, and vice versa for a decrease in vegetation. For a –10% change in urban fraction, the mean UHI intensity decreased the most in DJF nighttime (–0.19°C), followed by ASO nighttime (–0.13°C), DJF daytime (–0.11°C), and ASO daytime (–0.10°C). DJF nighttime had the highest mean UHI intensity in the control run (1.70°C), was the most sensitive to changes in urban fraction, and was the only case with a significant UHI intensity mean change for a 10% decrease in urban fraction. Therefore, increasing vegetation by a small amount could have major benefits.

Download

Share

COinS