Contents LAPS at AFWA GPS Meteorology LDAD DSS Gauge Effects LDAD QC Updates Etcetera Publications Contact The Editor Design: Wilfred von Dauster

Introduction As part of AWIPS, the Local Data Acquisition and Dissemination (LDAD) system is designed to acquire local datasets, perform quality control checks on the incoming data, and disseminate weather information–using advanced visualization and integration techniques–to decisionmakers in local and state communities and the public at large. LDAD will automate National Weather Service (NWS) field office interactions with local data observation systems, spotter networks, cooperative observers, and local emergency preparedness planners. Concurrent with the NWS modernization, state and local municipal governments, public utility companies, research organizations, and private industries have been installing meteorological observing systems suited to their respective needs. Local datasets slated for integration into AWIPS include hydrological observations from urban flood control districts, road weather observing networks, boundary layer profilers from urban air pollution monitoring systems, and airport wind shear alert systems. FSL developed the LDAD observation Quality Control and Monitoring System (QCMS) to check these datasets upon ingest into AWIPS. In this article we describe the NWS quality control requirements, summarize the initial QCMS capability included in the AWIPS Build 4.1 software version (see the November 1997 issue of the FSL Forum), describe subsequent upgrades made in Build 4.2, and discuss future upgrades. NWS Quality Control Requirements In 1994 the NWS developed requirements for the quality control of incoming data to the AWIPS system. This document, NWS Techniques Specification Package (TSP) 88-21-R2, describes the two categories of quality control checks–static and dynamic–for a variety of observation types, including surface, buoy, ship, profiler, aircraft, and rawinsonde data. The static quality control checks are single-station, single-time checks, which means that they are unaware of the previous and current meteorological or hydrologic situation described by other observations and grids. Checks in this category include validity, climatological, internal consistency, and vertical consistency. Although useful for locating extreme outliers in the observational database, the static checks have difficulty with statistically reasonable, but invalid data. To address these difficulties, the TSP document also describes dynamic checks, which refine the quality control information by taking advantage of other available hydrometeorological information. Examples of dynamic quality control checks include positional consistency, temporal consistency, spatial consistency, and model consistency. Another requirement covers the "data descriptor," a data structure intended to provide an overall opinion of the quality of an observation by combining the information from various quality control checks. Algorithms described to compute the data descriptor are a function of the types of quality control checks applied to the observation, the sophistication of those checks, and the departure of the observation from the expected values provided by the quality control checks. The NWS further requires that Weather Forecast Office (WFO) personnel be able to view and, if desired, override the results of the quality control procedures. They also require that a database be available for storing and logging those results, which must be provided to NOAA and non-NOAA staff in charge of station maintenance. QCMS Implementation for AWIPS Although time constraints prevented full implementation of the TSP requirements before the AWIPS deadlines, the following LDAD QCMS capabilities were completed in the first implementation. Note that only hourly surface and buoy observations of sea-level pressure, temperature, winds, and humidity were processed for this build. Quality Control Processing – Two quality control checks – one static and one dynamic – were used used in the initial QCMS version for LDAD. The checks were a validity check that compares observed values to specified tolerance limits and a spatial consistency check that compares observed values to estimated values derived using observations at neighboring locations. Both checks were implemented as part of the Mesoscale Analysis and Prediction System (MAPS) Surface Assimilation System, known as MSAS. The LDAD implementation of MSAS ingests, quality controls, and analyzes most AWIPS surface observations contained in a domain covering the 48 contiguous states and neighboring areas of Canada and Mexico. The observations include surface data available over the Satellite Broadcast Network, i.e., standard meteorological aviation reports (METARs), Profiler Surface Observing System (PSOS), and buoy station reports, as well as local mesonet reports available at each WFO. The MSAS spatial consistency, or "buddy" check is performed using an optimal interpolation (OI) technique. At each observation location, the difference between the value measured and the value analyzed by OI is computed. If the magnitude of the difference is small, the observation agrees with its neighbors and is considered correct. If the difference is large, however, either the observation being checked or one of the observations used in the analysis is bad. To determine which is the case, a reanalysis to the observation location is performed by eliminating one neighboring observation at a time. If successively eliminating each neighbor does not produce an analysis that agrees with the target observation (the one being checked), the observation is flagged as "bad." If eliminating one of the neighboring observations produces an analysis that agrees with the target observation, then the target observation is flagged as "good" and the neighbor is flagged as "suspect." Suspect observations are not used in subsequent OI analyses. To improve the OI performance, the MSAS analysis fields from the previous hour are used as background grids. The analyses provide an accurate 1-hour persistence forecast and allow the incorporation of previous surface observations, thus improving temporal continuity. The differences between the observations and the background are calculated and then interpolated to each observation point before the OI analysis is performed. Also, uniform distribution of the neighboring observations used in the spatial consistency check is guaranteed (when possible) by locating the nearest observation in each of eight directional sectors distributed around the target observation. Temperature observations are converted to potential temperature before application of the spatial consistency check. Potential temperature varies more smoothly over mountainous terrain when the boundary layer is relatively deep and well mixed, a marked advantage during daytime hours. For example, potential temperature gradients associated with fronts tend to be well defined during the day even in mountainous terrain. Unfortunately, this advantage often disappears at night when cool air pools in valleys. To improve the efficacy of the spatial consistency check in these circumstances, elevation differences are incorporated to help model the horizontal correlation between mountain stations. The error threshold (to which the absolute value of the difference between analyzed and observed values is compared) is a function of the expected OI analysis error, and takes into account the distance and elevation differences between the target observation and the OI analyzed observations. Station Monitoring – The LDAD QCMS was also designed to keep statistics on the frequency and magnitude of the observational errors encountered for NWS sea-level pressure, potential temperature, dewpoint, and surface wind. The statistics are calculated for all stations in the MSAS domain, and stations from different networks are kept statistically separate. Specifically, the stratifications maintained for AWIPS include: automated SOS (ASOS), Surface Aviation Observations (SAOs, defined as METAR manual), Auto (METAR automated, but not ASOS), Buoy, and NOAA Profiler Network (NPN). Local mesonets are stratified by provider, as in CDOT (the Colorado Department of Transportation). The statistics are reported using hourly, daily, weekly, and monthly quality control messages. The hourly messages provide the total number of observations for each variable, the number of observations that failed the quality control check, the station names for the failed observations, and the error and threshold values for each of the ailed observations. The error is defined as the difference between the quality control analysis value and the observed value, as computed in the spatial consistency check described above. The daily, weekly, and monthly messages list stations that have failed the quality control checks at least 25% of the reporting period, and also contain the percentage of failed observations and the average error and rms error for individual stations and for all stations combined. Subjective Intervention – Also introduced in the initial QCMS version were two text files, a "reject" and an "accept" list, which provide the capability to subjectively override the results of the quality control checks. The reject list is an inventory of stations and associated input observations that will be labeled as bad regardless of the outcome of the quality control checks. The accept list is the corresponding list for stations that will be labeled as good regardless of the outcome of the quality control. Applications reading the lists, such as the MSAS analysis (example shown in Figure 1), will then reject or accept the stations specified. In both cases, observations associated with the stations in the lists can be individually flagged. For example, wind observations at a particular station may be added to the reject list, but not temperature observations. Figure 1. An AWIPS screen showing the MSAS sea-level pressure analysis for 2200 UTC on 15 September 1999, overlayed on NOWRAD radar, and zoomed on Hurricane Floyd. The MSAS program runs locally on AWIPS and ingests (among other surface observations) LDAD mesonet observations quality controlled by the QCMS. Quality control and station monitoring procedures are not affected by subjective intervention lists, with the sole exception that observations on the reject list will be labeled as suspect and not used to check the spatial consistency of neighboring observations. This allows the WFO to continue monitoring the performance of the stations contained in the lists. For example, a hydrometeorological technician may notice a station with wind observations that fail the quality control checks a large percentage of the time, and choose to add that station to the reject list. However, once the observation failure rate at the station falls back to near zero (possibly due to an anemometer that has been repaired), the technician will likely delete that station from the list. Quality Control Output – The quality control messages described above are available to forecasters and other WFO personnel via the AWIPS text workstation, and to data providers via the LDAD dissemination system. Current QCMS Upgrades Upgrades made to the LDAD observation QCMS in Build 4.2 fell into three main categories: implementation of subhourly quality control processing, establishment of a quality control database, and introduction of AWIPS quality control displays. QC Processing – Table 1 lists the quality control checks applied to the various surface observations in AWIPS. Notice that additional meteorological variables have been added to the QCMS processing, along with the additional quality control checks and subhourly processing. The latter is important not only because many networks now report multiple times per hour, but also because predicting observation arrival times is difficult at best. With once-per-hour quality control processing, observations at stations reporting early in the hour may not be quality controlled for 15–20 minutes, while observations at stations reporting late may not be quality controlled at all. These problems were solved with a QCMS process that checks multiple times per hour for newly arrived observations. Observations not quality controlled previously are then immediately quality controlled with validity, temporal consistency, and internal consistency checks. Table 2 lists the tolerance limits for the current implementation of the validity and temporal consistency checks. Validity checks, as previously described, restrict each observation to falling within a TSP-specified set of tolerance limits. Temporal consistency checks restrict the temporal rate of change of each observation to a set of (other) TSP-specified tolerance limits. In both cases, observations not falling within the limits are flagged as failing the respective quality control check. Table 1 Quality control checks implemented in AWIPS for surface observations input through the LDAD and Satellite Broadcast Network communications networks. Table 2 QCMS tolerance limits for the validity and temporal consistency checks implemented in AWIPS. Observations not falling between these limits are flagged as bad. The QCMS internal consistency checks enforce reasonable meteorological relationships among observations measured at a single station. For example, a dewpoint temperature observation must not exceed the temperature observation made at the same station. If it does, both the dewpoint and temperature observation are flagged as failing the internal consistency check. Pressure internal consistency checks include a comparison of pressure change observations at each station with a difference of the current station pressure and the station pressure three hours earlier, and a comparison of the reported sea-level pressure with a sea-level pressure estimated from the station pressure and the 12-hour mean surface temperature. In the former check, if the reported 3-hour pressure change observation does not match the calculated observation, then only the reported observation is flagged as bad. In the latter check, however, if the reported sea-level pressure does not match the calculated observation, then both sea-level and station pressure observations are flagged as failing. The QCMS also now quality controls LDAD station pressure, altimeter setting, pressure change, relative humidity, visibility, and precipitation observations, in addition to sea-level pressure, temperature, wind, and dewpoint temperature observations. Because of time constraints, however, neither the station monitoring capabilities nor the quality control messages have yet been updated to include these new variables. Similarily, subhourly checks for Satellite Broadcast Network surface observations have not yet been implemented. Quality Control Database – The 4.2 version of the QCMS also introduced the LDAD netCDF files containing raw observations, and the following quality control structures: a "QC-applied" bitmap indicating which quality control checks were applied to each observation, a "QC results" bitmap indicating the results of the various quality control checks, and a "QC departures" array holding the estimated values calculated by the quality control checks (such as the analysis-minus-observation value calculated by the spatial consistency check). Also included in the netCDF files are single-character "data descriptors," the data structures intended to define an overall opinion of the quality of each observation by combining the information from the various quality control checks. In the initial system, an observation was labeled as bad if it failed either of the quality control checks, and good if it passed both. The QCMS is now upgraded to label the quality of observations as preliminary, coarse, screened, verified, questionable, or erroneous, based on the quality control checks applied to the observation and the sophistication of those checks. For example, observations that have passed their validity check are labeled as "coarse," indicating that coarse quality control checks have been applied and passed. If those same observations then undergo temporal and internal consistency checks and pass, they receive a "screened" descriptor, but if they fail either check they are labeled as questionable. Likewise, observations failing the validity check are labeled as "erroneous," and observations passing all checks (including the spatial consistency checks) are given a descriptor of "verified." Data descriptors of good and bad are used to indicate observations included in the accept and reject lists, respectively. Table 3 provides a complete list of the QCMS data descriptors. The netCDF files form the initial version of a quality control database that will allow AWIPS programs to quickly and easily retrieve data of a certain quality for inclusion in their applications. Quality Control Output – In addition to the text quality control messages, AWIPS Build 4.2 contained, for the first time, the ability to display LDAD quality control information along with the raw observations. The displays are part of the D2D component of AWIPS, an interactive display for two-dimensional data. The quality control displays consist of color-coded station plots. Stations with observations found bad by the QCMS are distinctly colored to indicate possible problems with their reported data (Figure 2). Pointing and clicking on any station invokes the display of a small quality control table indicating which quality control checks have been applied at the time of the display, which ones have been passed, and which ones have been failed. Plots are automatically updated as new data arrive and are quality controlled. Table 3 QCMS data descriptor definitions used in AWIPS. Stage 1 QC consists of observation validity checks; Stage 2 QC, temporal and internal consistency checks; Stage 3 QC, spatial consistency checks. Figure 2. AWIPS quality control display showing the results of the QCMS quality control checks applied to LDAD mesonet data. Stations with observations that passed the quality control are colored green; stations with observations found bad are distinctly colored to indicate possible problems with their reported data. The QC table shown indicates that the Remote Automated Weather System (RAWS) station at Black Rock, Arizona, passed the validity and internal consistency checks for all observations reported, but failed the temporal consistency check for both temperature and dewpoint temperature observations. The availability of the quality control messages through the AWIPS text workstation and the LDAD dissemination system has been retained. Future Upgrades More work will be required to meet the NWS requirements for the quality control of incoming data. Of highest priority is the continued implementation of the TSP specifications for the quality control of surface data. This involves the installation of additional quality control checks for the surface observations discussed here, as well as the implementation of quality control checks for other observation types. Many observation types have not yet been ingested or processed by the QCMS, including many observations that are of critical importance to the hydrological, oceanographic, and climatological communities. We need to upgrade QCMS station monitoring capabilities to include additional observation types. As the number of observations arriving in each WFO continues to grow, automated techniques designed to assist maintenance personnel in locating invalid data values and/or equipment failures become increasingly important. As such, graphical displays designed to better present and summarize the station monitoring statistics are being considered, possibly as part of the existing Web-based AWIPS monitoring system. Future work should also include enhancements to existing quality control checks. For example, current plans for MSAS include an increase in both spatial and temporal resolution. These upgrades will produce not only improved surface analyses but also subhourly results and improved background grids for the spatial consistency check. Finally, since quality control ingest into AWIPS applications and visualization programs requires the quality control database, there is a compelling need for extending it to include all AWIPS observations, not just LDAD surface mesonets. [Editor's Note: More information on this topic is available on the QCMS Website, including bibliographies.] ( Patricia A. Miller is Chief of the Scientific Applications Group of the Systems Development Division, headed by U. Herbert Grote)