Abstract
We revised the calibration of a water vapor Raman lidar by co-located radiosoundings for a site in the high European Arctic. For this purpose, we defined robust criteria for a valid calibration. One of these criteria is the logarithm of the water vapor mixing ratio between the sonde and the lidar. With an error analysis, we showed that for our site correlations smaller than 0.95 could be explained neither by noise in the lidar nor by wrong assumptions concerning the aerosol or Rayleigh extinction. However, highly variable correlation coefficients between sonde and consecutive lidar profiles were found, suggesting that small scale variability of the humidity was our largest source of error. Therefore, not all co-located radiosoundings are useful for lidar calibration. As we assumed these changes to be non-systematic, averaging over several independent measurements increased the calibration’s quality. The calibration of the water vapor measurements from the lidar for individual profiles varied by less than ±5%. The seasonal median, used for calibration in this study, was stable and reliable (confidence ±1% for the season with most calibration profiles). Thus, the water vapor mixing ratio profiles from the Koldewey Aerosol Raman Lidar (KARL) are very accurate. They show high temporal variability up to 4 km altitude and, therefore, provide additional, independent information to the radiosonde.
Highlights
In the Arctic, average temperatures rise twice as fast as on global average; the so-called “Arctic Amplification” of global warming
C · P is the lidar signal in the unit photons. (For Licel transient recorders, c might be the inverse of laser shots written in one data file.) In these units, the error of the lidar signal due to photon noise can be expressed by the standard deviation, which is the square root of the signal c · P
We found no seasonal drift in these profile specific calibration factors, and, it was valid to calculate the median over all correction factors at the specified heights in all suitable profiles (Ci, light green I in Figure 3) to determine one calibration factor per season
Summary
In the Arctic, average temperatures rise twice as fast as on global average; the so-called “Arctic Amplification” of global warming. To derive the humidity from the lidar signal, a calibration is necessary to link the ratio of the lidar profiles to the actual water vapor mixing ratio. This calibration can be done by either measuring the transmission of the different detection branches of the lidar or comparison of the lidar product with external instruments. Radiometer humidity profiles have high uncertainties [24] and the integrated water vapor, which is considered to be more reliable, cannot be used for calibration because KARL lacks exact measurement in the lowermost 350 m. By using a correlation coefficient between the mixing ratio profiles of radiosonde and lidar, it is shown that correlations below 0.95 could not be explained by noise in the data but indicate different meteorological conditions even for contemporaneous profiles
Sign-in/Register to view highlights & summary 
Published Version (
Free)
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call