A grid model of direct conversion between zenith tropospheric delay and precipitable water vapor in tropical regions.

Accurately measuring precipitable water vapor (PWV) is challenging, especially in the absence of the measured meteorological data. We propose a novel grid model, CZP, that directly converts zenith tropospheric delay to PWV, leveraging ERA5 reanalysis data from 2016 to 2019. The CZP model considers the seasonal variability of the PWV and the spatial characteristics of the conversion coefficients, achieving high accuracy PWV retrieval in the tropics efficiently. Our findings demonstrate the CZP model's high accuracy and reliability across tropical IGS stations, with a mean bias and root mean square (RMS) well within acceptable limits compared to both ERA5, GNSS and radiosonde (RS) PWV measurements. The bias and RMS values for CZP PWV compared to GNSS PWV are less than 1.1 mm and 1.4 mm, while those values with respect to RS PWV are 1.61 mm and 3.11 mm, respectively. Furthermore, the real-time CZP PWV estimates show superior accuracy over the GPT3 model, with mean bias and RMS of − 0.55 mm and 3.55 mm, respectively. Such results underscore the CZP model's potential for enhancing weather prediction and climate research in data-scarce environments. [ABSTRACT FROM AUTHOR]