DAQUAMAP: Data Quality Monitoring in MAP
- An Initiative to Assess the Performance of Ground Based Meteorological Observations in MAP

Introduction

DAQUAMAP is an initiative to make an assessment of the performance and quality of the meteorological surface measurements in the MAP high resolution data domain. This initiative has been started in order to improve the quality of the existing networks for the field phase, as it has been stated in the MAP Design Proposal. The project is funded as a part of the optional EUMETNET program MAP-NWS.

Motivation

In the MAP high resolution data domain, far more than 10000 meteorological stations are operated by several dozens of institutions. Usually each institution uses own procedures for the quality control and the investigation of representativity of their stations. It would be difficult to obtain a dataset of more or less homogeneous quality. Therefore it seems reasonable to undertake some overall QC (quality control) effort. The QC efforts used by the different network operators should be documented for the MAP data users. Hopefully, the project will stimulate the experience and idea exchange in the field of QC of meteorological data among the operators of meteorological stations in the MAP region.

The overall quality control could be performed with an operational analysis-assimilation procedure. But these algorithms are tailored for the use in the initialization process of a numerical weather prediction model. There, dynamic constraints are very important as are the extensive use of first guess fields, optimum interpolation and variational approaches, which do not always provide satisfactory results in complex terrain. In order to evaluate the three dimensional state of the atmosphere with the greatest fidelity possible a method which makes extensive use of higher order derivatives will be applied to check and correct the meteorological data (Pöttschacher et al. 1997).

From experience we know that the re-examination of observational data of a field experiment is a long-lasting process. In order to initialize necessary remedy actions before the field phase, QC efforts should be started as early as possible.

From this motivation the following aims for DAQUAMAP result:

1. Provide an overview of QC efforts and correction methods for MAP data users and stimulate exchange of experience within MAP countries.
2. Monitor the performance of pressure, temperature and humidity measurements at stations below 750 m asl with the 2D version of VERA's (VERA = Vienna enhanced resolution analysis; Pöttschacher et al. 1996) error detection module. Forward results to network operators and MDC.
3. Implement and apply to all stations a 3D version of VERA's error detection module.

The state of the project

Documentation of QC Efforts

To meet the first goal, a ÔWorkshop on Quality Control of Meteorological Data in MAP' was held in Vienna on January 26./27. of this year. 40 experts from 19 institutions in 10 MAP-countries attended this workshop. The presentations - a part will be available in workshop proceedings later on - provided an excellent overview of the different QC procedures. Besides that, this workshop yielded further significant results:

1. The existing MAP WGObs (Working Group of Observational Networks) was renamed to WG ROUND (Working Group of Routine Network Data). The WG will act as a network in order to achieve the objectives and fulfill the tasks. The department of Meteorology and Geophysics of the University of Vienna serves as the focal point. The core of the WG is formed by a principle contact person for each institution represented in the WG.
2. Since there will not be enough time to implement comparable QC procedures in all institutions before the GOP and the SOP, the DAQUAMAP procedure (Pöttschacher et al 1997) was accepted as common QC procedure for all data in the MDC. A first report with results for the GTS stations has been finished by IMGW and is being sent to all institutions which transmit data to the GTS. As a next step, all data (also non GTS) available on MDC will be checked. IMGW will accomplish and send out a second report covering the MAP season 1997 before the GOP starts. There was agreement that the stations operators should give information about their interpretation of DAQUAMAP results to the project team. Results of this QC procedure will be made accessible to the MAP community after confirmation with the station operators.
3. The present state of QC procedures and data flagging will be documented in a summary containing information about:
   - how data are controlled
   - how data are corrected
   - how missing data are treated
   - how data are homogenized
   - how or whether controlled/corrected/interpolated/homogenized data are flagged.

Monitoring of Data Quality

The results of the error detection and correction module of VERA for msl pressure, temperature and humidity the GTS stations from 1.1.1997 until 31.12.1997 are being examined currently. A first report containing a full description of the method applied and a discussion of the results will be sent out to the institutions operating GTS stations in the MAP high resolution data domain in the next days. Figure 1 shows an example of a summary.

Percentage of "Rough Errors" >=  9 hPa
as determined with VERA's Error Detection Module 1.1.97 - 31.12.97

Figure 1. Percentage of `rough errors' (`best estimate' for correction) > 9 hPa. Stations with 0% rough errors are represented by a dot.

References:

Pöttschacher, W., C. Häberli and R. Steinacker, 1997: The Error Detection and Correction Module of CALRAS and VERA. MAP Newsletter, 7, 88-89.
Pöttschacher, W., R. Steinacker and M. Dorninger, 1996: VERA - A high resolution analysis scheme for the atmosphere over complex terrain. MAP Newsletter, 5, 64-65