Research Work
Contents below present the results of my selected studies.
Any collaborations, suggestions, and comments are welcome!
Mobile location data have emerged as a pivotal asset for analyzing travelers' spatial behaviors and movement patterns. In the context of air travel, the data empower researchers to gain empirical insights into travelers' choices of airports. This study employs mobile location data to scrutinize the market shares and infer catchment areas of three primary hub airports within the New York Metropolitan Area. Our study, together with Teixeira and Derudder (2021), helps contribute to a better understanding of competitive airport dynamics in the New York Metropolitan Area. In addition, the mobile location data allow us to calibrate the two key components of the Huff Gravity Model, which is frequently used in existing studies focusing on airport competition and catchment areas in Multiple Airport Regions (MARs). Our investigation underscores that the application of the Huff Model should not follow a uniform approach across different scenarios. The dynamics of airport competition and ground access alternatives exhibit unique characteristics within each MAR. Furthermore, our study unveils inherent quality challenges associated with mobile location data. Future studies intending to incorporate mobile location data are advised to conduct preliminary assessments of data quality before embarking on empirical analyses.
​
Keywords: Airport choice, Mobile location data, Spatial analysis, Airport catchment area Huff model
We reviews 87 air travel demand studies published from 2010 to 2020 and summarizes these studies using their input data and primary analytical methods. We also devise and conduct three citation analyses to further explore the relationships among the reviewed studies. Our review finds that a typical empirical study of air travel demand analysis would focus on the demand at the national level, employ time-series data concerning socio-economic and airline operational factors and use time-series based methods to estimate the relationship among the selected time-series. These studies are mostly applying existing analytical frameworks to specific problems rather than developing original methods, therefore their relationship to each other is parallel rather than sequential. A small number of references are frequently cited by the reviewed studies primarily because of their methodological contribution to time-series analysis. A common limitation of existing literature is that very few reviewed studies provide validation of their analyses. In addition, methods that are not regression or time-series based have very limited application in this area so far, so are the non-convention data such as mobility data or social media data. Besides providing a systematic summary of recent publications in a specific field, this review uses a relatively objective and replicable framework to compare and link studies by their references, which can be visualized by the figures included in this review. This review is expected to benefit future researchers that are interested in either air transportation or the application of time-series forecasting in an applied domain.
​
Keywords: Literature review, Citation analysis, Time-series forecasting, Air travel demand
Using Hartsfield-Jackson Atlanta International Airport as a special case, this study proposes a theoretical framework for quantifying and comparing the overall cost of parking and Transportation Network Companies (TNCs) services such as Uber. Based on the cost comparison, we build a web application to visualize the utility advantage area and summarize the corresponding demographic information. Our study has the potential to benefit airports, TNC operators, and travelers. Using our app, these stakeholders can visualize and measure potential tradeoffs between parking and TNC Ridesharing services.
​
Keywords: Geographic Information System, Airport Ground Access, Transportation Network Companies, Spatial Analysis
​
Access the app via https://avationresearch.shinyapps.io/ATL_Parking_Uber/
​
The app will output the catchment area of airport parking and Uber based on the traveling inputs.
We focus on addressing what forces are moving the airline stock price during the pandemic? In this study, we use JETS, an Exchange Traded Fund (ETF) tracking major aviation stocks, as the proxy for stocks of US aviation-related businesses. Three time series, including the S&P500 index, daily COVID-19 confirmed cases in the US, and daily passenger security checkpoints throughput number provided by the US Transportation Security Administration, are selected as potential factors causing JETS price to fluctuate. A special Vector Autoregression (VAR) model based on a procedure introduced by Toda and Yamamoto (1995) is used to reveal the temporal causal effects between these time series. The analysis finds that two factors, daily COVID-19 confirmed cases and TSA throughput numbers, are significant in predicting the price movements of JETS during March – December 2020. A simulated trade strategy based on the findings of our analysis is found to be able to achieve substantial financial gain using the historical trading data, which to some extent confirms the validity and the implication of this study.
​
Keywords: Multivariate Time Series Analysis, Granger Causality, Stock, COVID-19 pandemic, Aviation Economics