|
The HRW product aims to provide detailed and frequently updated sets of winds in the daytime, from HRVIS Seviri data (nominal horizontal resolution of 1 km), as a provision of dynamical information from the SAFNWC/MSG. The product includes wind pressure level information, and wind quality control flagging giving some indication of error in probabilistic terms. This product should be useful in near real time applications as nowcasting (used in synergy with the diverse other data available to the forecaster) and mesoscale analysis or short range forecasting.
HRW output is similar to other products of Atmospheric Motion Vectors in the GTS: BUFR bulletins. Some processing parameters can be configured, in particular the region to be processed. A two scale procedure (24 pixel -basic scale- and 12 pixel -detailed scale-) is suggested as baseline. Detailed tracers are calculated in areas where no basic tracers can be found, or these are large suggesting the possibility of a more detailed search.
Six different steps take part in HRW algorithm, which are described next:
1. Image preprocessing: HRVIS Seviri Reflectances are normalized with the sun zenith angle, with data values ranging from 0 to 255 (8 bit data, to avoid too low contrasts).
2. Tracer determination: two methods are used, with the objective of finding as many tracers as possible all throughout the image.
Gradient or sharp edge: fast and efficient search of well defined cloud edges.
Tracer characteristics: fills holes of coverage, with a longer but still reasonable computing time. Each location is checked to find a brightness threshold (separating the cloudiness in front from the background), small scattering of IR channel temperatures in the target (to avoid multilevel cloudiness), and a well defined shape of the target bright pixels (to avoid too linear cloudy elements).
3. Tracer level assignation: two inputs are used: the IR10.8 channel brightness temperature and the output of SAFNWC/Cloud Type (PGE02) product, if available. A Cloud Type is defined for each tracer as the most common value of SAFNWC/CT output inside a square box around the tracer location. According to the Cloud Type value defined for the tracer, the level assignation is then made through a linear vertical and horizontal interpolation to the centre of the tracer of NWP forecast or climatological temperature profiles, considering:
The Cloud Top for tracers related to"High semitransparent thick clouds" and "High semitransparent meanly thick clouds", through the coldest non isolated class in the smoothed temperature histogram. .
The Cloud Base for tracers related to all other Cloud Types, calculated through TBASE = TMEAN +Ö2 sTEMP (J.Schmetz et al. 1996).
Tracers related to "Cloud free land or sea", "Land or sea contaminated by ice or snow", "Fractional clouds" and "High semitransparent thin clouds" are elliminated because of their incorrectness or bad validation statistics.
In case PGE02 output is not available, the level assignation is done for all clouds with the Cloud Base.
4. Tracking: given a set of tracers for a HRVIS region at "t-Dt slot", tracking areas of the same size are defined at "t slot" to match tracking candidates. To reduce the tracking area, it is defined through the extrapolation of linearly interpolated NWP forecast wind profiles, up to a configurable distance.
Two well known methods can be used for this process: Euclidean difference or Cross correlation. The best three tracking centres are kept, to perform a final selection step at the Quality Control process.
5. Quality Control: the Quality Indicator Method, developed at Eumetsat for AMV calculation at MPEF (K.Holmlund 1998), has been adapted for HRW product. Several consistency tests are if possible computed, considering:
Vector/direction/speed consistency with AMVs in the previous slot (Temporal test).
Vector consistency with neighbour AMVs (Spatial test).
Vector consistency with NWP forecast winds (Guess test).
Vector consistency with simultaneous Basic winds (Two scale test, only for Detailed winds).
An individual Quality Index (QI) is calculated for each Quality test, with the normalized statistical fitting functions described by the method. An overall Quality Index is calculated as the average sum of partial QI. A double contribution of spatial test is included.
6. Orographic Flag: an Orographic flag can also be calculated since HRW v2.0, related to the detection of land tracers, tracers blocked by orography and orographic waves. The algorithm calculates for each geographical box of 1x1 degree of latitude and longitude the medium and maximum box heights, and makes a barometric conversion to pressure with NWP data to calculate the medium and minimum box pressures. These data are compared with the tracer pressure to calculate the values of the Orographic flag:
1: The tracer is below the level with the mean representative pressure of its geographical box (a significant orography influence is expected).
2: The tracer is below the highest level defined to have any orographic influence: minimum representative pressure-25 hPa (less orography influence is expected).
Else, if vertical stability is found at the tracer location and speed is at least 5 m/s, previous positions of the tracer up to two hours are calculated with the corresponding AMV trying to detect an orographic wave. With these conditions, other possible orographic flag values are:
3: An obstacle has been found against the flux, with the tracer below the level defined to have any orographic influence at any of its previous positions (possible orographic wave).
4: No obstacle has been found, but stability condition is still present at all previous positions (the obstacle might be further).
In any other case, the value of Orographic flag is:
5: No orographic influence has been found.
7. AMV Selection: a Comparison Test is assigned to AMVs, when more than an AMV is available per tracer. These tests are used for final AMV Selection, where only one AMV per tracer can be kept. Six different tests are calculated:
From AMV calculation process: Tracking correlation test.
From Quality control process: Temporal, spatial, guess tests; Two scale test also for Detailed winds.
From Orographic flag process: Orographic test.
Possible test values are:
3: when the AMV is the best of all for the corresponding criterion.
2: when the AMV is slightly worse than others for the criterion.
1: when the AMV is rather worse than others for the criterion.
0: when the comparison flag could not be calculated for the AMV.
The selection criterion is based on the AMV with the best Comparison Tests. In case of draw, the AMV with best Guess Test is selected, or else the AMV with the best Tracking Correlation Test.
Product Configuration files: HRW Model Configuration and Region Configuration files.
Seviri data: Full Resolution MSG HRVIS & IR10.8 Seviri data for the working region and times "t" and "t-dt".
NWP data: Fields of Temperature and Rectangular components of the wind for as many as possible of the following pressure levels: 1000, 925, 850, 700, 500, 400, 300, 250, 200 hPa. Also surface temperature and Geopotential for pressure levels 1000, 925, 850, 700, 500 hPa in case Orographic flag is to be calculated.
SAFNWC/Cloud Type output file for the running slot and region.
List of tracers and predecessor winds for time "t-dt" (in case they are available).
The product resolution is defined by the user; the default value is 24 km for the basic wind dataset, and the half for the detailed wind dataset.
Two files for basic and detailed winds, or just one in the case of a single scale, in form of BUFR bulletins for each processed region is obtained with the following content (see SAF/NWC/INM/SW/ICD/3 for description of PGE09 BUFR structure).
Basic parameters of the wind:
Initial latitude and longitude of the tracer.
Latitude and longitude increments in the tracer displacement.
Temperature and pressure level at the initial tracer location.
Direction and speed of the wind.
Quality Index of the wind.
Orographic flag.
Cloud Type related to the tracer.
Processing indicators for the wind:
Type of tracer (only for the detailed HRW product file: fine scale related to a wide coarse scale tracer, other type of fine scale, coarse scale non wide tracer).
Pressure level assignment method.
Tracking method.
Test indicators: which of the partial QI consistency tests (two scale, temporal, spatial, forecast test) and orographic flag have been applied, and if the selected wind is best or slightly or rather worse than others for the considered test).
Wind number.
Wind number of the prior wind (if any, if option persistent tracers).
Nominal confidence threshold.
Number of winds computed for the tracer (1 to 3).
Number of levels available for the Temperature and Wind guess.
Tracer size.
General processing indicators:
Satellite and generating centre.
Generating application.
Satellite cycle of final image.
Year, month, day, hour, minute of final image.
Time increment from initial to final image.
Search zone increment at nadir.
Satellite instrument and generating method.
Channel used, central frequency and width.
Correlation method (plus gap applied yes or not).
Origin of first guess.
Quality control source, manual or automatic.
Two different model configuration files have been defined,
optimized for the running of the product in National or Continental areas. Next
figures show typical displays of HRW considering these areas, with the
recommended values indicated by these reference files: safnwc_pge09nat.cfm and
safnwc_pge09cont.cfm.
