Merging of the Stage II/IV Hourly Analyses
for Precipitation Assimilation
Since its implementation in July 2001, the EDAS precipitation assimilation uses the NCEP Stage II hourly precipitation analysis as the source of observed precipitation. In the new package, the NCEP Stage IV will be the primary source of the precipitation input, supplemented by the Stage II. Both analyses are produced at NCEP, and they are based on hourly radar and gauge observations.
The Stage IV analysis merges the regional multi-sensor precipitation analyses produced from the twelve River Forecast Centers (RFC) over the CONUS. It benefits from some manual quality control performed at the RFCs, so it is generally considered to be of higher quality than the Stage II product. It is however not as timely as the Stage II - the Stage II is produced at NCEP at approximately 40 minutes after the top of the hour. The timeliness of the Stage IV depends on the transmission of the regional analyses, it is generally a few hours delayed, and the latest available Stage IV analyses often have only partial coverage, since analyses from RFCs arrive at different times of the day.
1. Merging procedure
Prior to model integration for each of the 3-hour EDAS segment, we collect the available Stage II/IV analyses for these three hours to produce a merged analysis by using Stage IV values when available (unless the point is within NWRFC domain - see below), and fill in any gaps with available Stage II values. If there is no Stage IV data at all for that hour, then Stage II alone is used. If both analyses are not available, which is likely for the last hour of EDAS in operational runs due to time constraints, then the entire precipitation input field consists of missing points, and precipitation assimilation would not be done for the hour.
The analyses we receive from NWRFC often contain spurious precipitation "bull's eyes", and it appears that in this region the Stage IV is not as reliable as the Stage II. We decide to use an RFC domain mask during the Stage II/IV merging process so that in the NWRFC domain, Stage II is used instead of Stage IV.
An example of the merging is shown for the 12Z 20030112 cycle. The pre-forecast data assimilation period covers 00-12Z 12 Jan. During the first nine hours, the Stage IV data coverage is relatively complete. For the hour ending 10Z, coverage for Stage IV decreased and Stage II was used to fill in. For 11Z and 12Z there was no Stage IV at all, and Stage II analyses were used for those two hours.
2. Experiments Conducted
a) Parallel test in 32km Eta
The merging of the Stage II/Stage IV analyses was tested in the 32km 'ETAL' during 12Z 2002/11/19 - 12Z 2003/01/13. Prior to 00Z 2002/12/24, the Stage IV analysis in the NWRFC domain was not excluded during the data merge, so the spurious precipitation bull's eyes in the NWRFC region did get into the merged analyses, though in the subsequent gross QC check most of these bad values got filtered out.
Related web pages:
b) Parallel test in 12km Eta
Starting 00Z 2002/12/28, the merging of the Stage II/Stage IV analyses for EDAS is tested along with other components of upgrades (for the upcoming Spring 2003 Eta/EDAS implementation) is tested in the 12km Eta.
Related web pages:
During the first five weeks of testing on the 32km Eta, the contribution from NWRFC was not masked out from the Stage IV analyses, so the precipitation bull's eyes in the NWRFC domain got into the merged analyses. Most of them were filtered out by the gross error check before the merged product was used in the assimilation.
The most significant impact of introducing the Stage IV analysis into precipitation assimilation is the improvement of model precipitation field during EDAS, as shown in the EDAS precipitation scores (blue: cloud top assimilation run; red: control run). Since the model's soil moisture field is driven by the amount of model precipitation during EDAS, this means the soil moisture field would be improved as a result. When we sum up Stage II/IV analyses into 24h (12Z-12Z) totals and compare them to the CPC 1/8 deg gauge-based daily analysis (which is the most reliable measure of daily analysis available), we find that sometimes Stage IV is clearly better than Stage II, other times Stage II is better than Stage IV, often they are comparable but overall Stage IV appears to have an edge over Stage II.
There is a small positive impact on the precipitation forecast. Impact on upper air fields is minimal.
The merging of Stage II/Stage IV analyses for EDAS precipitation assimilation is scheduled for operational implementation in Spring 2003.