Corrections of Humidity Measurement Errors from the Vaisala RS80 Radiosonde—Application to TOGA COARE Data

A series of laboratory tests have been conducted on several different batches of Vaisala RS80 radiosondes to understand and develop methods to correct six humidity measurement errors, including chemical contamination, temperature dependence, basic calibration model, ground check, sensor aging, and sensor arm heating.The contamination and temperature-dependence (TD) errors dominate total errors.The chemical contamination error produces a dry bias, and is due to the occupation of binding sites in the sensor polymer by nonwater molecules emitted from the sonde packaging material.The magnitude of the dry bias depends on sensor polymer type (RS80-A and RS80-H), age of the sonde, relative humidity (RH), and temperature, and it exists throughout the troposphere.The contamination error generally increases with age and RH, and is larger for the RS80-H than the RS80-A.It is ϳ2% and ϳ10% at saturation for 1-yr-old RS80-A and RS80-H sondes, respectively.The TD error for the RS80-A results from an approximation of a linear function of temperature to the actual nonlinear temperature dependence of the sensor, and also introduces a dry bias.The TD error mainly exists at temperatures below Ϫ20ЊC, increases substantially with decreasing temperatures below Ϫ30ЊC, and is much larger for the RS80-A than the RS80-H.The RS80-A's TD correction (C TA ) dominates the total correction at temperatures below Ϫ40ЊC and has a correction factor [C TA ϭ (RH) (C TA factor)] of 0.15, 0.75, and 2.3 at Ϫ40Њ, Ϫ60Њ, and Ϫ80ЊC, respectively.The correction methods are applied to 8129 Vaisala RS80 soundings collected during the Tropical Ocean and Global Atmosphere (TOGA) Coupled Ocean-Atmosphere Response Experiment (COARE) and are applicable to RS80 radiosonde data from other field experiments and historical and operational radiosonde datasets.The methods are validated by examining various summary plots of the TOGA COARE data and comparing them with other independent data.The corrections greatly improve the accuracy of the TOGA COARE radiosonde dataset.These correction methods have their own uncertainties and may not correct all errors in Vaisala RS80 humidity data.Analyses of these uncertainties are presented in the paper.