Earth and Space Science paper Ferreira et al. 2026 published

The total mass of atmospheric water vapor over a unit of area, or precipitable water vapor (PWV), is an important parameter for predicting our weather and climate. PWV values based on balloon-borne radiosonde observations of pressure, temperature, and relative humidity have served as a benchmark for remote sensing PWV measurements and are key to climatological studies on PWV. This study compares alternative PWV algorithms and derives formulas for error propagation. Usual approximations are found to yield a slight, latitude-dependent underestimation of PWV, while  traditional radiosonde reports introduce an additional negative bias due to the encoding of air temperature. Besides radiosonde biases, our error propagation analysis indicates that PWV uncertainty  is dominated by the uncertainty in temperature and humidity measurements associated with unit-to-unit variability for the same radiosonde type. Random errors occurring in soundings have a much  smaller influence and are negligible with high vertical resolution data. The set of error propagation formulas developed here is expected to improve the assessment of remote sensing PWV techniques, as well as the estimation of PWV trends based on radiosonde. Additionally, we suggest considering the use of GPS altitude instead of pressure to improve the accuracy of PWV  calculations.