| Author: | Tan, Xiaoyue |
| Title: | Generation of high-quality daily nighttime light time series through uncertainty modelling and cloud removal |
| Advisors: | Zhu, Xiaolin (LSGI) |
| Degree: | Ph.D. |
| Year: | 2023 |
| Subject: | Remote sensing Image processing -- Digital techniques Hong Kong Polytechnic University -- Dissertations |
| Department: | Department of Land Surveying and Geo-Informatics |
| Pages: | 167 pages : color illustrations |
| Language: | English |
| Abstract: | Remotely sensed light imagery offers a unique perspective depicting human activities. Since the 1970s, several satellite sensors have been launched to detect nighttime lights (NTL). The released NTL datasets have yielded valuable insights into various fields and applications. Recently, continuous global daily NTL products released by NASA's Black Marble nighttime lights product suite (VNP46) are increasingly utilized due to their unique capabilities in capturing rapid human, such as cultural patterns on holidays, before and after power outages, and during in COVID-19 pandemic, as well as dynamics in natural systems including natural disasters, snow cover, nocturnal fire, and air conditions. However, the daily NTL observation from satellites is susceptible to various factors (e.g., acquisition conditions, observational geometry, surface environments, and cloud covers) that lead to undesired noise and invalid observations, thus limiting its application, especially for daily NTL time series. This thesis endeavors to generate high-quality daily NTL time series by addressing the following key questions: (1) What are the spatiotemporal patterns of uncertainties in daily satellite nighttime light time series? (2) How does the satellite viewing angle, as a unique uncertainty in daily NTL observation, affect satellite-observed NTL radiance? (3) How can we achieve accurate predictions for cloud-contaminated pixels in NTL images and reconstruct high-quality daily NTL time series? To address question 1, I investigated the spatiotemporal patterns and drivers of uncertainties in daily NTL time series. This research employed NASA’s Black Marble products, the lunar-BRDF corrected daily NTL data, which represents the most advanced and promising product with broad applications. To better understand the uncertainty across multiple spatial and temporal scales, this study proposed a spatial-temporal hierarchical analysis strategy to separate uncertainties from different sources, and evaluated the effects of multiple factors on daily NTL time series. The experiments conducted on two populous regions show that: (1) The daily NTL in Northern America has variations up to 50% of the annual average, which is stronger than in East Asia, with variations up to 25%. (2) the dominant source of daily NTL uncertainty is different in the two regions: Seasonal variations dominate the NTL variations in Northern America, while day-to-day changing factors dominate daily NTL variations in East Asia. (3) Environmental factors and observational conditions show spatially varying impacts on NTL. Specifically, aerosol exhibit opposite impacts on rural and urban areas; significant impacts of moonlight are mainly distributed in rural areas; the impact of satellite viewing angle is less pronounced and frequent in East Asian cities compared to those in North America. This research revealed the essential knowledge about variations of satellite-derived NTL and will benefit the processing and utilization of daily NTL products. To address Question 2, special attention was given to the uncertainty caused by the angular effect, which is a unique source of uncertainty in daily NTL observations when compared to monthly or annually composited datasets. Generally, angular observations lead to inconsistencies among observations over the same area on different days, introducing uncertainty into daily NTL time series. In this study, first, I proposed a conceptual model of the angular effect and hypothesized the mechanism of how urban 3D landscapes form the anisotropic characteristics of artificial light observations. Second, I quantified the spatial patterns of the angular effect within five representative cities, and identified three distinctive types of angular effects: negative, U-shaped, and positive. Subsequently, the contribution of landscape factors to the direction (i.e., the type) and magnitude (i.e., NTL change rate with angle) of the angular effect is quantified using multinomial logistic regression and mediation analyses, respectively. The results show that the direction of the angular effect is mainly controlled by building height which determines the blocked and visible parts of artificial light at different satellite viewing angles. The magnitude of the angular effect is determined by both NTL brightness and landscape factors. The mediation analysis shows that landscape factors can have a direct effect on the magnitude of the angular effect as well as an indirect effect on the magnitude by affecting NTL brightness. Among the landscape factors, both vegetation and buildings are indicated to be significantly influential factors with direct and indirect effects. The findings of this study deepen our understanding of the NTL angular effect, and help us better monitor high-frequency socioeconomic activities. To address Question 3, I developed an effective method, named as Cloud Removing bY Synergizing spatio-TemporAL information (CRYSTAL), to generate cloud-free NTL images with satisfactorily accurate pixel brightness and spatial continuity. The development of cloud removal algorithms is crucial as data gaps resulting from cloud contamination and low-quality observations inevitably hinder the effectiveness of these applications. Although a temporal gap-filling method is employed in recent Black Marble NTL products to produce seamless images, the filled images are unsuitable for quantitative analysis. Simulation experiments show that CRYSTAL can produce more accurate results than the temporal gap-filling method in fifteen cities worldwide, with an average RMSE reduction of 33.69%. Images generated by CRYSTAL restore temporal variances in NTL and are thus suitable for multi-temporal quantitative analysis. CRYSTAL can reconstruct daily NTL time series by filling gaps using available partially clear images. Experiments in two cities demonstrated that the reconstructed time series had 31.85% more valid values than the original time series and effectively revealed urban dynamics during the early stages of the coronavirus disease 2019 pandemic. In summary, CRYSTAL is a novel and effective gap-filling method for the restoration of invalid NTL observations and has the potential to generate high-quality NTL data for use in future applications. In summary, in this thesis, I (1) investigated the spatiotemporal patterns and underlying factors that contribute to uncertainties in daily NTL time series, (2) explored the mechanism and quantifies the uncertainties associated with the angular effect, and (3) developed an effective cloud removal method for NTL images, which enables the generation of high-quality daily NTL time series. Overall, this thesis sheds light on the generation of high-quality daily NTL time series through the comprehensive modeling of uncertainties and the development of an innovative cloud removal method. |
| Rights: | All rights reserved |
| Access: | open access |
Copyright Undertaking
As a bona fide Library user, I declare that:
- I will abide by the rules and legal ordinances governing copyright regarding the use of the Database.
- I will use the Database for the purpose of my research or private study only and not for circulation or further reproduction or any other purpose.
- I agree to indemnify and hold the University harmless from and against any loss, damage, cost, liability or expenses arising from copyright infringement or unauthorized usage.
By downloading any item(s) listed above, you acknowledge that you have read and understood the copyright undertaking as stated above, and agree to be bound by all of its terms.
Please use this identifier to cite or link to this item:
https://theses.lib.polyu.edu.hk/handle/200/12742