Radio frequency (RF) signals are used in Global Navigation Satellite Systems (GNSS) for positioning applications, however, it can also be used to monitor the atmosphere. RF signals can be affected by changes in the atmospheric refractivity index along their propagation path. This change of refractive index along the path of the signal in the troposphere causes a delay to the signal known as the Tropospheric Delay (TD). The TD in the zenith direction (ZTD) has already been used to derive the amount of precipitable water at a given site because the refractive index of air in the atmosphere is proportional to the environmental variables: temperature (T), pressure (P) and water vapour partial pressure (e). However, other environmental variables such as T have not been derived from the ZTD. Thus, this thesis presents a novel algorithm to estimate temperature from GNSS data for monitoring urban heat island intensity (UHII).
An urban heat island (UHI) occurs when an urban area is warmer than its adjacent rural areas. It exacerbates heat waves, leading to increased energy consumption and adverse effects to the environment and to human health. UHIs are monitored using remote sensing techniques, which allow the monitoring of large geographical areas with low time resolution. However, the study of UHIs within a city requires better spatial and temporal resolution. It is also desired to monitor the UHI in real-time. The algorithm presented in this thesis allows UHI monitoring with higher spatial and temporal resolution using a GNSS network.
The algorithm developed in this research has 6 inputs: the thickness of the troposphere, air pressure, water vapor partial pressure and the vertical profile of the refractive index obtained with radiosonde data. Another input is the ZTD obtained from the Precise Point Positioning (PPP) technique. The algorithm solves for temperature at the point where the GNSS data was collected. To validate the output of the algorithm, estimated T at 5 locations at 00:00UTC and 12:00 UTC have been compared to values of T from meteorological data near the GNSS station at the same times. Hourly data for 20 days in year 2017 has been used. An average difference of less than 1 ºC has been found for data collected during the summer.
In order to measure the intensity of the UHI, it is necessary to measure the temperature at two locations simultaneously: an urban and an adjacent rural location nearby. The algorithm has been tested and validated using two publicly available datasets containing daily GNSS and meteorological data from Los Angeles, California (LA), USA and Hong Kong Special Administrative Region, China (HK). Also, the algorithm has been tested with an experimentally collected dataset containing hourly GNSS and meteorological data from Ningbo, China (NB). It has been found that an UHI with an intensity of 3.5 ºC existed in LA during the winter 2017.The UHI detected in HK during the summer 2017 had an intensity of 4 ºC and in NB had an intensity of 2 ºC.
|Date of Award||8 Jul 2020|
- Univerisity of Nottingham
|Supervisor||Lawrence Lau (Supervisor), Yu-Ting Tang (Supervisor) & Terry Moore (Supervisor)|
- GNSS remote sensing
- UHI monitoring
- Tropospheric delay