Decisions, Decisions, Decisions

By Tom Frame:

In September of this year I was fortunate enough to escape my desk for a while and participate in the first leg of the observational field campaign for the DIAMET project.
The main aim of DIAMET is to improve the understanding of the contribution of diabatic process to the development and structure of extra-tropical cyclones, through both modelling and observation. The observational component makes heavy use of the FAAM aircraft to fly into frontal regions and make in-situ observations using the various instruments mounted on the hull as well as dropping sondes.

I did get to fly, which is pretty exhilarating – particularly at 100ft over the ocean! However most of the time on the field campaign is spent deciding why, when and where the aircraft should fly. As a researcher used to looking at forecasts after they have been verified, this planning process itself is always an interesting experience of using forecast data in real time.

During the field campaign each day starts with a 9:00 o’clock meeting here in Reading to discuss the latest forecasts, followed by a 9:30 conference call including all collaborators from other institutions as well as representatives from FAAM. During this call we have to decide whether to stake a claim on the aircraft against other projects who might want to use it. We must make our initial expression of interest two days ahead and make a final decision 24 hours prior to flying.

The decision is challenging as it must take into account many factors and constraints:

  1. SCIENCE – Do the forecasts show weather of sufficient scientific interest to justify a flight?
  2. TIME –  We can only fly between 9:00am and 5:30pm Monday to Friday for a three week period.
  3. SPACE – The aircraft has limited range – see radar image below with a typical flight track overlaid. We cannot drop sondes over land (except Ireland apparently!). There are many areas where maneuvers are restricted by other air traffic or we must obtain permission in advance before entering. We cannot enter French airspace.
  4. MONEY – We have a limited number of flight hours and dropsondes to use. Do we use resources on the present case and run the risk of running out when an even better case comes along or do we let this case go and run the risk that nothing comes along later?
  5. UNCERTAINTY – Do we believe given the forecast that the nature, location and timing of the weather feature we’re interested in are sufficiently certain to occur in the way we hope?

As a general rule weather forecasting uncertainty propagates up from small space and time scales to larger scales with increasing lead time. This means that when a decision can be made depends on the lead time at which the uncertainty is deemed small enough relative to the constraints on SCIENCE, TIME, SPACE and MONEY.

A good example is the cases of Monday 19th and Tuesday 20th September 2011. By Friday the 16th September, the synoptic situation for Monday and Tuesday was certain: Large region of relative high pressure over Europe, a Large low pressure system over Iceland with long trailing cold front (See the forecast below).

The scientific interest was also clear as there was a very high probability of frontal waves developing on the trailing cold front – pretty much every ensemble member in the Met Office and ECMWF forecast were showing waves.  But here is where the issue of the scale of the constraints versus the uncertainty kicks in. At what time will the frontal wave be in a suitable location for the plane, will it fit in with our 9:00 to 5:30 Monday to Friday flying times? Below the forecast fronts from each Met Office ensemble member are plotted for the Monday and Tuesday – cold front in blue (green for control forecast), warm front in red (yellow for control forecast) – compare the scale of the spread to typical flight track of the plane above. We estimate a higher chance of a frontal wave reaching the Northern end of the front on Tuesday than Monday – but on Tuesday it’s possible the front itself has passed into the dreaded French airspace by mid-day leaving almost no flying time.

In the end it was decided to have a conference call on Sunday to make a final decision – sacrificing a little weekend time in exchange for making the best decision. The result – the flight took place on the Tuesday and successfully observed a frontal wave.

Leave a Reply

Your email address will not be published. Required fields are marked *