Back to results list
Please use this identifier to cite or link to this item:
|Title:||Remote sensing of atmospheric water vapor field with tomography using multi-sensor data||Authors:||Chen, Biyan||Advisors:||Chen, Wu (LSGI)||Keywords:||Water vapor, Atmospheric -- Remote sensing.
Atmosphere -- Remote sensing.
|Issue Date:||2017||Publisher:||The Hong Kong Polytechnic University||Abstract:||Water vapor is an important component in the atmosphere. Due to its significant role in the transfer of energy in hydrological processes, water vapor has a profound influence on climate and weather. Monitoring the changes in atmospheric water vapor content can not only detect global and regional climate change, but also predict severe weather systems. Moreover, water vapor can retard the propagation of microwave signals passing through the atmosphere. The water vapor induced path delay is one of the limiting factors in making high-accuracy range measurements for space geodetic systems such as global navigation satellite system (GNSS). Accurate knowledge of the water vapor distribution in the atmosphere is therefore of great significane for these studies and applications. Water vapor tomography is a dedicated method to resolve the three dimensional (3D) distribution of water vapor with high accuracy and reliability. The main objective of this thesis is to develop new methods of water vapor tomography by using multi-sensor water vapor observations. In water vapor tomographic modeling, the troposphere is usually discretized into a number of voxels. To determine the water vapor field, each voxel requires to be crossed by a certain number of GNSS rays from various directions. During a tomographic modeling process, however, some voxels are usually unable to be crossed by any rays, leading to a rank-deficient problem to the tomographic modeling. To maximize the number of voxels with GNSS ray crossings, a voxel-optimized tomographic approach is developed in this thesis. This optimization method consists of four components: 1) top boundary determination, 2) vertical layer discretization, 3) horizontal boundary optimization, and 4) selection of optimal vertical and horizontal resolutions. The optimal discretization of tomographic model of Hong Kong region is investigated towards both high accuracy and high spatial resolution of the solutions. Based on the 40-year radiosonde profiles, the top boundary of the tomographic model in Hong Kong is determined as 8.5 km since it is found that water vapor above 8.5 km can be negligible. The horizontal boundary optimization is achieved by moving the voxel location in latitude and longitude until the maximum number of voxels with ray crossings is reached. For the vertical layer height optimization, the layer thickness increases with altitude considering the atmospheric physical property that the water vapor content decreases with the altitude. Based on extensive experiments, an optimal discretization of the tomographic model for Hong Kong region is determined. The tropospheric path delay can be divided into a hydrostatic part and a wet part, where the latter one is exploited in the water vapor tomography. To estimate the tropospheric wet delay, the hydrostatic delay should be first accurately determined. Usually, the hydrostatic delay is calculated by empirical models with the use of meteorological parameters. A comprehensive evaluation of the performances of 9 zenith hydrostatic delay (ZHD) and 18 zenith wet delay (ZWD) models in China was performed using the radiosonde data as references. For GNSS meteorology, Baby ZHD model is the best model with an accuracy of 6.0 mm. If only the pressure data are available, Saastamoinen model is a proper and the only choice to calculate the ZHD.
To assess the quality of water vapor data from various observation systems, an intercomparison study was conducted for water vapor data derived from GPS, radiosonde, water vapor radiometer (WVR), non-hydrostatic model (NHM), and European Center for Medium-Range Weather Forecasts (ECMWF). For ZWD comparison with radiosonde data, ECMWF achieves the highest accuracy of 17.73 mm (~2.87 mm in precipitable water vapor (PWV)). GPS, WVR, and NHM have RMS errors of 18.06 mm (~2.93 mm in PWV), 18.15 mm (~2.94 mm in PWV), and 29.53 mm (~4.78 mm in PWV), respectively. Slant wet delays (SWD) estimated by GPS were assessed by SWDs derived from ECMWF, an overall accuracy of 36.44 mm (~5.90 mm in slant PWV) is yielded. Water vapor tomographic experiments were carried out using multiple data from GPS, radiosonde, WVR, NHM, sunphototmeter, and synoptic observations in Hong Kong. Experimental results have revealed that the best vertical constraint scheme is using average radiosonde profiles observed during the three days prior to the tomographic epoch. In the evaluation by radiosonde observations, the multi-sensor tomographic wet refractivity fields achieved an overall accuracy of 7.13 mm/km. In the vertical direction, RMS errors generally decrease with altitude from 11.44 mm/km at the lowest layer (0 to 0.4 km) to 3.30 mm/km at the uppermost layer (7.5 to 8.5 km). The tomographic results obtain RMS errors in the range of 6~9 mm/km at the horizontal grids when compared with ECMWF data. An important goal of water vapor tomography is to benefit the extreme weather prediction and thus to mitigate ensuing hazards. Due to the transfer of energy in the atmospheric processes, atmospheric water vapor has a strong influence on formation and lifecycle of severe weathers. Three heavy precipitation events that occurred in Hong Kong were investigated to examine the potential of water vapor tomography in extreme weather prediction. Several positive findings demonstrated the ability of tomography in forecasting heavy rains as it can detect atmospheric instability before the events.
|Description:||PolyU Library Call No.: [THS] LG51 .H577P LSGI 2017 Chen
xviii, 214 pages :color illustrations
|URI:||http://hdl.handle.net/10397/67213||Rights:||All rights reserved.|
|Appears in Collections:||Thesis|
Show full item record
Files in This Item:
|b29527764_link.htm||For PolyU Users||208 B||HTML||View/Open|
|b29527764_ira.pdf||For All Users (Non-printable)||10.58 MB||Adobe PDF||View/Open|
Citations as of Jun 18, 2018
Citations as of Jun 18, 2018
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.