Laserfiche WebLink
<br />I <br />I <br />I <br />I <br />I <br />I <br />I <br />I <br />I <br />I <br />I <br />I <br />I <br />I <br />I <br />I <br />I <br />I <br />I <br /> <br />noted when the radiometer was changed from stationary to mobile operation, or <br /> <br />during the reverse procedure. Following the field season it was found that a <br /> <br /> <br />power supply in the 20 GHz channel was faulty, and was intermittently changing <br /> <br />its voltage output. During the season the data were adjusted for these voltage <br /> <br /> <br />fluctuations by using different tipping curve calibrations. Provided the voltage <br /> <br /> <br />was steady once it shifted to a new level, the data can likely be adjusted using <br /> <br /> <br />different tipping curve factors, but it will also be necessary to use the USBR <br /> <br />radiometer vapor values to verify the correctness of the adjusted values. The <br /> <br />affected periods will require further editing which may result in some loss of <br /> <br />data. <br /> <br />Radiometer calibration will follow the procedure of Decker and Schroeder <br /> <br />(1991) for determining tipping curve calibration factors. This procedure relies <br /> <br />on the fact that in a horizontally stratified atmosphere absorption varies <br /> <br />linearly with air mass. Measurements of brightness temperature are taken at <br /> <br />several elevation angles (air masses). Brightness temperatures are converted to <br /> <br />.absorption through the Rayleigh-Jeans approximation. The measurements are then <br /> <br />adjusted through the use of an appropriate set of calibration factors to produce <br /> <br />a linear relationship between absorption and air mass, that passes through the <br /> <br />origin (zero air mass). This ensures that the downwelling radiation is being <br /> <br />correctly measured. The factors adjust for losses occurring between internal <br /> <br />components such as switches and reference loads, and correct for circuitry <br /> <br />changes that can develop over time. Numerous tipping curve calibrations were <br /> <br />performed during cloudless conditions in January, February and March 1991. <br /> <br />The computation of liquid and vapor values from the caUbrated measurements <br /> <br /> <br />of brightness temperature is made using the statistical retrieval technique <br /> <br /> <br />described by Hogg et a1. (1983). Atmospheric retrieval coefficients and the mean <br /> <br />19 <br /> <br />