Hurricane Harvey evolution process monitoring based on PWV

In 2017, Hurricane Harvey made two consecutive landfalls along the Gulf Coast of the United States on August 25th and August 30th, resulting in extreme precipitation and widespread flooding. The impacts were devastating, causing the loss of over 80 lives and significant economic damage. To understand the evolution of Hurricane Harvey, we utilized data from densely distributed permanent GPS stations for precipitable water vapour (PWV) and rainfall tracking from the Tropical Rainfall Measuring Mission (TRMM) 3B42 product, examining spatial and temporal characteristics. Applying the empirical orthogonal function (EOF) method, we analyzed typical spatial patterns of PWV before and after Harvey's passage, revealing an overall increasing trend in atmospheric moisture at various sites across the study area. This trend was particularly pronounced in regions affected by the first and second landfalls. By comparing the time series of PWV with the distance between GPS stations and the hurricane's eye, we calculated their correlation, finding a negative relationship, especially when the hurricane is gradually approaching the GPS stations, the correlation coefficient can reach -0.6 or a higher value. Notably, when the distance between GPS stations and the hurricane's eye was approximately 1000 km, PWV rapidly increased. As the water vapour values rose to around 55mm and maintained an upward trend, sustained precipitation occurred. During the passage of the hurricane, PWV remained at elevated levels, resulting in maximum rainfall. Once the hurricane moved away, PWV values rapidly decreased.