Measurement Uncertainties of Observations of Temperature and Water Vapor within GRUAN

The uncertainty of measurements of upper air essential climate variables, in particular temperature and water vapor in the upper troposphere and lower stratosphere limit the understanding of long term changes, but also of some details of physical and dynamical processes. Operational, data collection, and processing routines within the GCOS Reference Upper Air Network (GRUAN) aim to characterize all sources of measurement uncertainties and to quantify vertically-resolved systematic and random uncertainties. A detailed understanding of the measurement uncertainties enables the network evaluating the impact that changes in the measurement system have on the ability to detect long term trends and to evaluate the observations in the context of process studies. Long term climate observations are most strongly impacted by systematic changes, such as changes in calibration, changes in the correction of systematic errors or changes in the sensor response. Many systematic effects change from altitude layer to altitude layer and depend on the parameter that is being investigated. Manufacturing processes, instrumental changes and processing routines strive to minimize the impact of systematic effects on long term observations. A strong natural variability of tropospheric temperature and water vapor allows for instruments with larger random uncertainty. The much lower variability of stratospheric water vapor and lower sampling frequency requires instruments with lower random uncertainty. Here we discuss the current understanding of the sources of uncertainty for measurements of temperature in the troposphere and stratosphere using the Vaisala RS92 radiosonde and for water vapor in the troposphere and stratosphere using both the Cryogenic Frost point Hygrometer (CFH) as well as the Vaisala RS92.