A Procedure for Correcting Radiosonde Reports for Radiation Errors

Larry McMillin National Oceanic and Atmospheric Administration, National Environmental Satellite, Data, and Information Service, Satellite Research Laboratory, Washington, D.C.

Search for other papers by Larry McMillin in
Current site
Google Scholar
PubMed
Close
,
Michael Uddstrom National Oceanic and Atmospheric Administration, National Environmental Satellite, Data, and Information Service, Satellite Research Laboratory, Washington, D.C.

Search for other papers by Michael Uddstrom in
Current site
Google Scholar
PubMed
Close
, and
Alessandro Coletti S M System and Research Corporation, Landover, Maryland

Search for other papers by Alessandro Coletti in
Current site
Google Scholar
PubMed
Close
Full access

Abstract

Temperature sensors on radiosondes measure a temperature that is a balance between the temperature of the air and the temperature of the radiation environment of the sensor. Because of this balance, the temperature reported by a radiosonde differs from the true air temperature by an amount that is determined by the heat-transfer coefficient, the longwave emissivity of the sensor, the shortwave emissivity, the longwave flux on the sensor surface, the shortwave flux on the sensor surface, and the sensor temperature. Of these quantities, the heat-transfer coefficient is determined by properties of both the sensor and the atmosphere, the reflectivities are determined by the sensor, and the fluxes and air temperature are determined by the atmosphere. For a typical radiosonde, the radiative properties of the sensor can be determined, the coefficient of heat transfer can be estimated, and models exist for calculating the shortwave flux. In this paper, the authors show that a modification of the Elsasser formulation for infrared fluxes can be used to calculate the infrared flux. This provides sufficient information to solve for the temperature difference between the temperature sensor and the air. The method is used to calculate errors for some typical meteorological conditions for the white-coated VIZ sensor, made by VIZ Manufacturing Co.

The method was used to examine the range of radiation errors for typical conditions. Although the shortwave radiation error is generally recognized because it is observed in the day-to-night differences, it is demonstrated that the longwave radiation errors are significant and variable. The longwave error for the VIZ instrument can reach 3 K at 10 hPa for a winter profile and can exceed that when there is a stratospheric warming. While the longwave radiation error is generally considered to be a cooling effect, at 100 hPa the longwave radiation is a source of heating in a tropical atmosphere because the tropopause at 100 hPa is cold relative to the rest of the atmosphere that is radiating to the sensor. Clouds have significant effects on both the longwave and shortwave components. A cloud at the tropopause of a tropical atmosphere can change the error of a radiosonde above the cloud from a heating error to a cooling error. At sunrise and sunset, the change in the shortwave error is abrupt. Extremely accurate knowledge of the time and location of the radiosonde is required to reduce the uncertainty in temperature to less than 1 K at the upper levels at these times.

Abstract

Temperature sensors on radiosondes measure a temperature that is a balance between the temperature of the air and the temperature of the radiation environment of the sensor. Because of this balance, the temperature reported by a radiosonde differs from the true air temperature by an amount that is determined by the heat-transfer coefficient, the longwave emissivity of the sensor, the shortwave emissivity, the longwave flux on the sensor surface, the shortwave flux on the sensor surface, and the sensor temperature. Of these quantities, the heat-transfer coefficient is determined by properties of both the sensor and the atmosphere, the reflectivities are determined by the sensor, and the fluxes and air temperature are determined by the atmosphere. For a typical radiosonde, the radiative properties of the sensor can be determined, the coefficient of heat transfer can be estimated, and models exist for calculating the shortwave flux. In this paper, the authors show that a modification of the Elsasser formulation for infrared fluxes can be used to calculate the infrared flux. This provides sufficient information to solve for the temperature difference between the temperature sensor and the air. The method is used to calculate errors for some typical meteorological conditions for the white-coated VIZ sensor, made by VIZ Manufacturing Co.

The method was used to examine the range of radiation errors for typical conditions. Although the shortwave radiation error is generally recognized because it is observed in the day-to-night differences, it is demonstrated that the longwave radiation errors are significant and variable. The longwave error for the VIZ instrument can reach 3 K at 10 hPa for a winter profile and can exceed that when there is a stratospheric warming. While the longwave radiation error is generally considered to be a cooling effect, at 100 hPa the longwave radiation is a source of heating in a tropical atmosphere because the tropopause at 100 hPa is cold relative to the rest of the atmosphere that is radiating to the sensor. Clouds have significant effects on both the longwave and shortwave components. A cloud at the tropopause of a tropical atmosphere can change the error of a radiosonde above the cloud from a heating error to a cooling error. At sunrise and sunset, the change in the shortwave error is abrupt. Extremely accurate knowledge of the time and location of the radiosonde is required to reduce the uncertainty in temperature to less than 1 K at the upper levels at these times.

Save