METEOSAT rapid scan during MAP-SOP

MAP need for high frequency METEOSAT data

METEOSAT operational schedule ensures a reasonable spatial and temporal coverage of cloud systems within the MAP area, but a more flexible and powerful observing strategy is needed to meet MAP high-resolution data requirements. Needs for a rapid scan strategy during MAP can be seen in several aspects deriving from the MAP Science Plan (Bougeault 1998).

Deep convection

Purdom (1996a,b) has discussed the benefits of very rapid scan imagery (1 min interval) from GOES-8 for the study of explosive convection and demonstrated the great impact of rapid scan down to 30 sec repetition time for the analysis of the genesis and evolution of squall lines and their fine structure. Projects such as VORTEX (Rasmussen et al. 1994) have benefited from such imagery. Levizzani and Setv‡k (1996) have found striking reflectivity patterns on top of deep convective storms in Europe related to the injection of very small ice crystals into the higher troposphere and lower stratosphere from the storm top. Rapid scan observations conducted from GOES-8/9 over the US Great Plains document the time evolution of such objects demonstrating that an internal mechanism of storm evolution is responsible for their formation above the storms anvil.

Winds

Tests using high-resolution image sequences from GOES-8 at 30 sec, 1 and 3 min image repetition time were conducted on severe storms and hurricanes, as well as more common situations like winter storms with multiple cloud layers and trade wind flow over the ocean (Purdom 1996c). Results show that high-frequency imagery is very good for deriving cloud motion, even in the most complex situations, and improve cloud drift winds. In Fig.1 an example of cloud drift winds obtained at different image repetition times is proposed. High-frequency METEOSAT imagery over the MAP area can certainly contribute to wind derivation tests for ingestion into forecast models. Experiments on cloud drift winds will act as a benchmark for operational wind derivation schemes by the METEOSAT Product Extraction Facility (MPEF).

Rainfall estimation

MAP provides a great opportunity to test satellite rainfall estimation schemes given the amount of non-conventional meteorological data that is made available. High-frequency scanning strategies in turn allow for obtaining rainfall maps more often than the actual 30 min interval thus approaching the radar frequency. Better statistics and calibration results are then achievable. The structure of frontal rainbands in organized convection is another argument in favor of a higher scanning frequency. Small scale structures, though at METEOSAT spatial resolution at MAPs latitudes, will be closely followed and compared with model outputs.

Use for mesoscale analysis

Check of cloudiness model output is a rather appealing subject for MAP. High-resolution mesoscale models will benefit from the availability of cloudiness and cloud temperature maps at a few minute interval. High resolution models are already at a stage where additional data, if not only for verification purposes, are very beneficial on the very short range. The availability of frequent scans in the water vapor channel is also considered important for the analysis of convergence/divergence upper air structures. PV streamers research and upper level feature studies, in general, can also benefit from such images.

MAP as a natural laboratory for MPEF

Last, but not least, MAP provides the possibility for EUMETSAT and MPEF to take advantage of a very complex set of meteorological data normally not available within the METEOSAT field of view. The opportunity of using the MAP data setto test hypotheses, algorithms, scanning protocols, and the derivation of new meteorological products is very appealing. Strategies for METEOSAT Second Generation (MSG) and the derivation of products within EUMETSATs Nowcasting Satellite Application Facility (SAF) are to be considered as an important occasion for the forecaster community. High-frequency imagery plays in this sense an important role: its application to nowcasting product extraction should be taken into consideration. This would eventually lead to a better understanding of satellite features, normally observed at 30 min repetition time, but that can be better understood increasing the scanning frequency.

Proposed scanning strategy

In view of the above outlined scientific and operational needs a scanning strategy is proposed. The present proposal is by no means complete and the necessary checks have to be performed as to technical feasibility, cost analysis (both currently under consideration by EUMETSAT OPS), and other issues that might arise from future interactions between MAP scientists and EUMETSAT boards.

The following are the key features of the proposal:

1. Time period: August to November 1999.
2. METEOSAT 7, at that time the operational meteorological satellite, will perform its normal 30 min scanning schedule to ensure operational imagery and products to member states NWS.
3. METEOSAT 6, the back-up satellite parked 10 degrees W, will ensure rapid scan of the MAP area.
4. The time of rapid scanning will be requested by the user: i.e. the MAP Operation Center (MOC), will notify EUMETSAT OPS that potentially interesting weather features are developing and then a rapid scan is to be planned.
5. The lead time for the rapid scan mode request (see 4) will be at least 24 hours.
6. Scannings during the rapid scan day will cover the northern hemisphere and be interleaved with rapid scanning of the MAP area.
7. The rapid scan area is critically dependent on the number of rapid scans per half hour. In Fig.2 two examples of areas at 6 and 8 scans per hour are shown.
8. Rapid scan will consist of up to eight mini-scans per half-hour.
9. A maximum of six hours rapid scanning per day is deemed possible.
10. If no rapid scanning is requested the satellite will be reverted to default 30 min fullglobe scanning.

Other salient features of this scanning strategy are:

• Images have to be calibrated and fully navigated in all three channels.
• Full resolution visible imagery would be considered important.
• The rapid scan should always retain the nominal METEOSAT spatial resolution.

Special considerations have to be dedicated to EUMETSAT OPS processing chain for the additional load of data provided by METEOSAT 6. The issue of archiving and data transfer between EUMETSAT and MAP Data Center is a key point not yet considered. The storage space required for the amount of data required seems relatively high, but not impossible to handle. In summary it is deemed that the possibility exists for operating the proposed schedule and that all the details can be discussed and solutions found for the benefit of both MAP and EUMETSAT.

	Figure 1. Cloud drift winds with different time repetition scans derived during severe thunderstorms over Texas on 12 April, 1996. Left to right and top to bottom: 1) 1 km resolution visible image at 22:39:34 UTC; 2) 30 min interval; 3) 15 min; 4) 5 min; 5) 1 min; 6) 30 sec. (courtesy Purdom 1996c
	Figure 2. Area covered by METEOSAT 6 ra- pid scan mode: 6 scans per half hour (left) and 8 scans per half hour (right)

Bougeault, P. (Ed.), 1998: The MAP Science Plan. MAP Programme Office, under final discussion.
Levizzani, V., and M. Setv‡k, 1996: Multispectral, high-resolution satellite observations of plumes on top of convective storms. J. Atmos. Sci., 53, 361-369.
Purdom, J.F.W., 1996a: One minute interval imaging of atmospheric phenomena using NOAAs new generation of geostationary satellites. Prepr. Eight Conf. on Satellite Meteorology and Oceanography, AMS, 164-167.
-, 1996b: Advanced atmospheric studies using GOES-8/9 multichannel imagery. Proc. 1996 Meteorological Satellite Data Users Conference, EUMETSAT EUM P 19, 77-86.
-, 1996c: Detailed cloud motions from satellite imagery taken at thirty second, one and three minute intervals. Proc. Third Int. Winds Workshop, EUMETSAT EUM P 18, 137-146.
Rasmussen, E.N., J.M. Straka, R. Davies-Jones, C.A. Doswell III, F.H. Carr, M.D. Eilts, And D.R. McGorman, 1994: Verification of the Origin of Tornadoes Experiment (VORTEX). Bull. Amer. Meteorol. Soc., 75, 995-1006.