DAQUAMAP Activities

Christian Häberli, Dept. for Meteorology and Geophysics, 1090 Vienna, Austria

The project DAQUAMAP (Data quality control in MAP) was finished in the first half of 2003. The results are available from the MAP database at:
http://www.map.meteoswiss.ch/map_app/q_sopSurface_form.jsp?sdate=07.09.1999&edate=16.11.1999.

Users can retrieve:

Flags for measurements which violated a 'global range check'. This global range check consists of a comparison of each value from non-GTS network data against the maximum + threshold and minimum - threshold. The extreme values were determined from the GTS network in the whole domain. The threshold is 10 hPa for pressure, 10 K for potential temperature and the humidity measure (θ - θe), 10 m/s for the wind components.
Flags for humidity values where the dew point is higher than the temperature.
For each suspicious value the deviation is stored in the database. Suspicious values are those where the deviation for the station under consideration exceeds the upper quartile + 3 * interquartile range or lower quartile - 3 * interquartile range.
The median value for the MAP SOP of all deviations for pressure, temperature, humidity and the wind components for each station. Here we assume, that the whole SOP contains a mixture of weather types and therefore the provided values reflect the systematic.

Based on these results users of the MAP database can:

Exclude erroneous and/or suspicious values.
Assess the magnitude of suspicious values.
Remove the systematic error from each individual measurement.

Users will find all further information and explanations at the internet link given above.

A technical final report is in the review with the DAQUAMAP review panel, which is chaired by Hans Volkert. This report will be submitted to the MAP steering committee and then be published on the internet. A scientific final report was submitted as a publication in a reviewed journal.

DAQUAMAP II: Radiosonde Humidity Data Correction

The method and some results for the correction of Vaisala RS80 radiosonde humidity measurements were presented at the MAP meeting in Brig (Häberli et al., 2003). One version of the correction takes into account outliers, the temperature dependence error, the time lag error, the dry bias due to chemical contamination (which is a function of sonde age and relative humidity), the sensor arm heating error, the ground check and the sensor aging error. A second version treats all errors beside outliers with one single empirical function. The impact on the correction based on results for the version 1 algorithm on precipitable water is shown in Fig. 1. The mean age of the radiosondes flown at Cagliari (16560, upper panel) was 3.75 years compared to 0.5 year at Palma de Mallorca (08301, lower panel). This explains the much stronger impact of the correction for the Cagliari data than for Palma. The `diurnal cycle' visible in the upper panel is due to the fact that for noon soundings another production batch of radiosondes was used than for the other observation terms: The noon ascents were performed with approximately 1.3 year old sondes whereas at all other times the sondes had an age of almost 4 years.

For version 1 of the correction sounding meta data like sonde serial number, ground check data and pre-launch reference measurements are required. It was possible to collect this type of data for 10 stations in the MAP high resolution data area. The second version of the correction does not require any meta data. Therefore it is possible to perform the calculations for about 40 stations in the MAP high resolution data area and in its close vicinity.

Currently the calculations are being performed and the MAP database administrators are evaluating how these data can be made accessible to the users in the most convenient way. A publication describing the algorithms, the results and the impact is in preparation.

References

Häberli, C. (2003): Assessment and Correction of the Dry Bias In Radiosonde Humidity Data during the MAP-SOP. Int. Conf. Alpine Meteorology and MAP-Meeting 2003. May 19-23, 2003, Brig, Switzerland. Extended Abstracts Vol. A. Publications of MeteoSwiss. 66, 221-224.