WINTER STORM RECONNAISSANCE PROCEDURES FOR SDM TASK:
BY 1 pm prepare a
a) Flight request for a flight centered around 0000 UTC the second day (observing time).
b) Outlook (yes or no) for flight(s) centered around 0000 UTC the third daySTEPS:
1) Compile a comprehensive list of verification events based on HPC and other possible requests. Review request list from previous day and incorporate requests that are still valid but are missing from today's requests. Eliminate duplicate requests and keep only verification events that are 1-4 days in the future from observing time for west coast and Alaska cases and 2-5 days for east coast cases.2) Review forecast for next 6-7 days and specifically for each requested event:
PQPF forecast at http://wwwt.emc.ncep.noaa.gov/gmb/ens/index.html (Opens in new window.)
MSLP forecast at http://www.emc.ncep.noaa.gov/gmb/tpm/emchpc/ (Note to Operator - MSLP forecast is bookmarked under "NCO Job Related" -> "EMC/HPC",or simply click on above link to open each in a new window. )
Comparison between NCEP/CMC's QPF/PQPF at http://wwwt.emc.ncep.noaa.gov/gmb/yzhu/html/CPQPF_cmc.html3) Inspect the flight request and outlook information from previous day at:
http://wwwt.emc.ncep.noaa.gov/gmb/targobs/target/wsr2007.html. Check if flight(s) were requested and if so whether the corresponding verification events are still relevant (threatening weather still forecast, and/or similar cases are still being requested). In the unlikely event the forecast drastically changed from previous day and no threatening event is in the latest forecast, cancel the flight requested previous day by contacting CARCAH as soon as you made a decision, any time before take-off. Inspect the WSR outlook information from previous day and make sure that the same or similar verification cases are included in today's list unless their lead time is too short to be considered, or due to a major change in the forecasts a high impact event became much weaker or disappeared in the recent forecast. Consult with HPC if necessary to confirm any changes compared to previous day's list.4) Identify possible observing time(s) for each verification event. Use 1-3 day time window for west coast/Alaska cases and 2-4 days for east coast events. Check the number of total observing cases (which is typically higher than number of verification cases since often more than one observing time is used for each verification event in search of best observing time).
5) Run setup_wsr and enter each observing case.
6) Ask the operator to run the operational job. Once the operator finishes the job (it takes about 5-10mins), run wsr_grds.sh script under sdm account. This will put plots on the web and also print them.
7) Analyze the results, first look at summary charts (along with flilght bar charts). A good target has:
a) Large numerical values on summary and bar chart, given observartion - verification time window (the longer the time window, larger the values are). Compare with values from 2000 and 2001 experiments using top figures on page reached through "Reasearch Links, Sharan Majumdar" from Targeted Observations page at: http://wwwt.emc.ncep.noaa.gov/gmb/targobs/index.html
b) Relatively focused maximum
c) Maximum area that can be sampled by a flight
Next, look at signal variance evolution charts. Try to assess if signal around flight track would reach and amplify into verification area.
If there are more than one verification events targeted at same observing time, consider whether there is a common track that can sample both or all. When choosing a common track, weight your choice by the priority of cases considered.
Decide whether a flight can be selected that can symmetrically sample the central area of the senstivie region. We should request flights only if most indications are positive, suggesting a successful mission. All missions will be verified. If in doubt, request a flight and make note in flight request for CARCAH that we may cancel next day. Put note in flight request for next day SDM to check on the case and cancel if necessary.
8) Look for possible signs of spurious sensitivity:
i) When looking for sensitive area at different observing times for same verification case, sensitive area should approach verification region as observing time approaches verification time. If not, consider sensitive area spurious. Inspecting multiple observing times therefore can enhance confidence in results
ii) Sensitive area too far from verification region - given observing time - is another sign of spurious sensitivity. Average speed data impact travels is 30 degree longitude per day, can be as high as 45 degree per day
iii) Signal variance evolution chart suggests no signal reaches verificaiton region or it originates from region other than what we would sample
iv) 250 hPa height ensemble spread evolution suggest spread at observation time around flight path would not reach verification region by berification time
9) Run wrapup_wsr. Enter flight request (yes or no), flight number(s) and associated verification events. Enter OUTLOOK information. This is not a request but only a heads up that can be changed next day without any problem. Enter necessary comments if any for CARCAH and next day SDM on separate lines.