6.1 Land-based surface observations

Table 11 displays a summary of the views available when the options “surface” and “availability” are selected in the WDQMS web tool (for near-real-time NWP monitoring module and Monitoring category, respectively).  The maps display the availability performance for each station highlighting differences in reporting performances as well as issues related to the station metadata recorded in OSCAR/Surface. Regarding the reporting performances we have the following categories: not reporting during the period (black); reporting below expectation (orange and red); and fully reporting (green). The metadata issues flagged are the following: station ID not known to OSCAR/Surface (yellow); station with observations not expected during the period according to the schedule recorded in OSCAR/Surface (grey); station reporting more than expected (pink). Clicking on a dot on the map opens up a screen with detailed information about the station data availability including the access to a chart containing the availability statistics over time (30 days for 6-hourly and dally aggregations and 24 months for monthly aggregations), namely time series with the total number of received observations during the selected interval/variable as well as the number of expected observations according to OSCAR/Surface or the GBON draft provisions (options in the “Baseline” menu). A link to the station metadata in OSCAR/Surface is provided for all stations except for the ones displayed as yellow dots as their IDs (referred as TSI) are not recognised by OSCAR/Surface.  

Table 11 - This table summarises the content of WDQMS web tool views when monitoring the availability of land surface observations in the near-real-time NWP module is selected in the NWP module (i.e., Near-real-time NWP: land-surface observations ; Monitoring category: availability).

Table 12 provides a summary of the menu options available in the WDQMS web tool when “surface” is selected under the near-real-time NWP monitoring module and “timeliness”  under the Monitoring category. By selecting the date, time, monitoring centre and variable, a map will display station-level timeliness performance, highlighting differences in reporting according to thresholds defined for high-resolution NWP applications (see Section 3.2): 15 minutes, 30 minutes, and 2 hours.  Clicking on a dot on the map opens a screen with detailed information about the station's timeliness, including access to a chart containing the timeliness statistics over time (30 days for 6-hourly and daily aggregations and 24 months for monthly aggregations) calculated for the selected interval/variable.  It is worth noting that the 6-hourly intervals for a particular monitoring centre will allow one to see the timeliness details of individual observations that contribute to the average value displayed on the map. Additionally, when selecting the ‘all’ option for monitoring centres in the time series, the values for all centres are shown, with outliers highlighted by a red circle. Although negative values are not included in the calculation of the average, the number of negative counts is displayed in the time series pop-up.

Table 12 - This table summarises the content of WDQMS web tool views when monitoring the timeliness of land surface observations in the near-real-time NWP module is selected in the NWP module (i.e., Near-real-time NWP: land-surface observations; Monitoring category: timeliness).

The options available in the WDQMS web tool control menu when we select to monitor the “quality” (as Monitoring category) of “surface” (in the near-real-time NWP monitoring module) are summarised in Table 13. By selecting the date, time, monitoring centre and variable, a map will show how the observing network complies with the NWP requirements set for the selected variable (see section 3.3). The map shows the stations whose observations are considered useful for NWP purposes (greenish dots), and highlights major anomalies regarding observations that deviate from the model forecasts quite significantly (orange and red dots) and exceed the limits set for “Threshold” (Table 7).  Clicking on a dot on the map opens a screen with detailed information about the station data quality over time (30 days for 6-hourly and dally aggregations and 24 months for monthly aggregations), namely time series with O-B calculated for the selected interval/variable. It is worth noting that the 6-hourly intervals for a particular monitoring centre will allow to see the details of individual observations that contribute to the average value displayed on the map. This means that it is possible to check the O-B value, the usage (Status, i.e., used or not used because it was rejected by/before the assimilation) and the type of report (Type, i.e., TAC or BUFR) of a particular observation (identified by date and time) by hovering the mouse over each dot in the time series. In the monthly aggregations, both trueness and precision are integrated into a single accuracy metric (root mean square error of O-B), while in the 6-hourly and daily aggregations only trueness (average of O-B) is available. Also, the percentage of gross errors for the month is calculated in addition to the root mean square error. The “Alert” is an option in the temporal aggregation based on 5-day moving averages across the WIGOS Monitoring Centres designed to warn WMO RWCs about potential quality issues. This measure takes into account the persistence of an existing daily problem.

Table 13 - This table summarises the content of WDQMS web tool views when the options to monitor the quality of  surface-land observations are selected (i.e., near-real-time NWP monitoring module: surface; Monitoring category: quality).