Repeatability and measurement uncertainty of the United States meteorological rocketsonde

Abstract

Remote temperature measurements from satellites are currently validated by using rocketsonde and radiosonde data. The reliability of the validation is directly dependent upon how well these measurements and those from the satellites represent the atmosphere. Past experience indicates that remote and in situ mean temperatures usually agree within a few degrees, while agreement between individual observations is not as good. It has been suggested that this may be due to atmospheric variability observed by the rocketsonde and radiosonde techniques but unresolved by the satellite. This observed variability, especially when measurements are obtained close in time and space, is of particular concern since this raises the question whether the variability is due to natural atmospheric change or arises from instrumental instabilities. It is shown that the U.S. rocketsonde, the Super Loki Datasonde, provides reliable temperature measurements to about 1°C. A sample of measurements obtained close together in time is analyzed to provide this estimate. It is important that the Middle Atmosphere Program (MAP) take advantage of the rocketsonde's measurement capability not only to validate remote measurements but to obtain detailed measurements of localized phenomena in the middle atmosphere.