The manner in which a tropical storm forecast is created is a simple step by step procedure combining several diagnostic elements. Knowing the there is a simple process to follow is the easy part, actually interpreting the data that is at the disposal of the forecaster is where things get difficult. Forecasting a tropical storm is definitely a science, but that does not mean that it doesn’t help to have a little Merlin the Magician like juju at work.
Once a tropical depression has been identified the timetable for forecasts is every 6 hours. The exact process and tools at the disposal of each individual forecast team will usually vary slightly from one warning center to another. In the US, NHC is the agency the public is most familiar with, but each primary center always has alternate that backs them up in case some situation may arise in which a primary center cannot execute their duties. The military has generally filled this void, with KGWC (Global Weather Central) acting as either the alternate or conduit to transfer information.
With the mundane portion removed, a forecast for a tropical storm employs some or all of the following diagnostic tools. Not all tools available will always be used based on the preference of the lead forecaster standing shift. In some cases they may be used but with weighted values to skew toward those which the forecaster has a personal preference for.
* Satellite Imagery - Satellite photos primarily transmitted via the DMSP (Defense Meteorological Satellite Program) are available every 6 hours for nearly every portion of the globe, and when they are not, birds can be tasked to cover them.
* Satellite Fixes - Satellite Analysts (SA) can take satellite imagery and create what is called a “Fix.” A Fix is a position for the eye of system. Only a scant twenty years ago this was done by eye, but eventually systems like the short lived MIDDAS (Meteorological Data Display Access System) came into being and laid the groundwork for a new wave of computer assisted analysis. Between the knowledge of the SA and computer assistance reading thermal imagery, a TPPA can be made which helps define the system for manual/computer plotting so that that the size can be reflected on a chart along with the swath edge.
* Synoptic Data - Synoptic data takes all the land, air and sea reports and places them on a chart. The main things being looked at are wind speed, wind direction, and barometric pressure. Synoptic data will generally consist of all the following on the 850mb (Surface) chart: Land observations, ship reports (on and off time), buoy reports, upper air (UJ) readings from weather balloons, TPPA fixes if available. On the 200mb chart (upper air) the data will consist of pireps (pilot reports), upper air (UJ) readings, and satellite winds. These are generally done no less than every 12 hours.
* Computer assisted models (Aides) - There are usually a dozen or more (minimum) computer generated forecasts which will take the current fix position and create a full life span predicted track. Each aide is geared toward a specific scenario. An example would be an aide that always predicts a system that has a “recurve” path, while another will predict a cookie cutter straight path, and another a stair step path. On their own, few are ever fully accurate, but by combining their strengths in a hybrid forecast model there is a slightly more objective path often defined.
* Various atmospheric charts - There are always plenty of charts on hand that are computer generated to reflect the atmosphere at varying levels from 850mb - 200mb in order to create a multi-dimensional look at the atmosphere.
* Synoptic Analysis - This is when the synoptic data mentioned above is analyzed. Streamline analysis allows for systems to be defined so that a forecaster can see the interaction between each and even help get the first ideas of where a new cyclone may be appearing.
* Experience - As valuable as everything above is in formulating a forecast, experience plays a big part. There is true value in being able to have worked a couple hundred or even few hundred systems for the salty forecaster. When you have seen past systems there are times you can mentally identify similarities to new systems and use that knowledge to help in the present.
Every experience is helpful whether it be noting a late season anomaly in storms that form in a certain portion of a basin, or how storms develop that take over 72 hours to go from a TCFA (Tropical Cyclone Formation Alert) to a Tropical Depression (TD). There is a reason meteorologists are called “weather guessers” and the discipline motto is “Your guess is as good as mine.”
When all these fields are downloaded, plotted, analyzed, and the forecast team has weighed in with their individual observations (If requested) on each area they specialize in, the lead forecaster will put out a forecast that projects 72 hours forward - although in some cases it will extend to 96 except in Japan where 24 hours is their maximum projection for typhoons. Even though the forecast will be re-issued in 6 hours, the product (forecast) is used by other disciplines as well like OTSR’s (Optimum Track Ship Routers) that help keep everything on the water clear of a system, and to government and civilian customers. Having as much of an idea of what is expected over 72 hours as possible helps provide everyone the best odds of being properly prepared from an operational and safety standpoint.
Once the forecast has been finalized, it is transmitted internally to be merged with a document that has PLA’s (Plain Language Addresses) to make sure it gets to the proper customers outside of the immediate operational area. From there, if military of government assets are listed, it will go through a secure communications channel the encodes the traffic properly to help scramble recipients so that security for said assets can be maintained.
At that point, the forecaster begins writing his/her prognostic reasoning which details in usually two brief paragraphs how the forecast was arrived upon. It may state this reasoning in some detail or be somewhat vague - usually depending on the time demands and confidence of the forecaster in their product. As soon as that is completed, the process begins all over again preparing for the next forecast.
In a nutshell, that is how a tropical storm forecast is created. All of the data in the world is great, but never underestimate the role that experience and sometimes dumb luck is at play in creating an accurate forecast. It is a science and in many ways an art that requires a great memory, highly analytic mind, and the ability to make a decision knowing full well you may be horrifically wrong. That is just the experience of the author who has better than 350 systems under her belt and although considered excellent, was wrong far more than she would care to admit.