Logo Leibniz Universität Hannover
Logo: Institute of Meteorology and Climatology /Leibniz Universität Hannover
Logo Leibniz Universität Hannover
Logo: Institute of Meteorology and Climatology /Leibniz Universität Hannover
  • Zielgruppen
  • Suche

DIVMET - Flight trajectory planning in case of adverse weather

Author: Manuela Sauer

Flight trajectory planning in case of adverse weather

Adverse weather affects safety and efficiency of air traffic. Pilots usually avoid adverse weather regions to ensure safety for passengers and prevent the aircraft from any structural damages. Avoided weather situations are, among others, icing regions and severe turbulence areas. Sometimes the latter are indistinguishable from calm air masses because it occurs under clear air conditions (Clear Air Turbulence, CAT). More often though turbulence is visible by clouds in which up- and downdrafts are spatially close. Especially in convective clouds such as cumulus or cumulonimbus turbulence is strong. These clouds frequently develop into storms and thunderclouds, which can, if there are multiple clouds, unite in intensive cells or clusters. In such areas often heavy precipitation in form of rain, graupel or hail, lightning, fall winds and icing occurs along with strong turbulence. A circumnavigation of thunderstorms is unavoidable with regard to several factors.



Effects of thunderstorms and icing on air traffic participants

Turbulence can cause structural damage of the aircraft as well as cargo loaded, and lead to discomfort and risk of injury to passengers. The accumulation of ice results in structural changes and a concomitant change in aerodynamics and flight performance. In particular, icing impairs control-relevant measuring instruments. In these cases, it is the pilots task to interpret the instruments output correctly and to react appropriately. A well-known case is the AirFrance flight (AF447) from Rio de Janeiro to Paris on 31 May/1 June 2009, which ended tragically in a crash over the Atlantic due to an incorrect interpretation and therefore inappropriate response of the pilot.

In the following it is focused on the circumnavigation of thunderstorms, as many other factors among turbulence and icing are relevant as well.


The avoidance/diversion behavior of pilots in reality

To initiate any circumnavigation the pilot must meet various requirements:

  • First, he must be aware of the weather situation ahead and its potential dangers. Apart from a weather briefing prior to the flight, the pilot has only limited access to weather information en route. Besides the visual appearance gained by looking out the cockpit windows during daylight, the pilot has the aircrafts on-board radar, which provides information during day and nighttime. The on-board radar gives a picture of backscatter signals that represent the distribution of precipitation particles ahead. Storms are indicated by a particularly strong signal. Range and opening angle of such a radar device are limited to about 80 Nautical Miles (NM), max. 180 NM, and 120°. The appearance of partial shadow effects, which appear when cells in the vicinity backscatter the entire signal or intercept a low backscatter signal, is one disadvantage of radar devices. Some aircraft are additionally equipped with a so-called stormscope that can detect electric charge in the environment and thus suggest potential hazards of a storm, or at least intense convection and wind shear in the vicinity.
  • When circumnavigating a thunderstorm, international regulations state to avoid the cell by defined distances. The UK’s air traffic control NATS recommends avoiding the storm by ten to 20 NM depending on the flight level. The Federal Aviation Administration (FAA) differs in the same range according to the intensity of the storm. However, the final decision whether there is a deviation from the planned trajectory and which distance is maintained, is up to the pilot.
  • A further factor to consider is the coordination with the appropriate air traffic control (ATC). Unlike evasive maneuvers due to other traffic, a deviation forced by weather is mostly considered by the pilot himself. Any deviation from the planned route must be announced and approved by ATC in terms of the surrounding traffic. Each change of course also needs to be communicated. In order to keep frequency of communication as low as possible, the number of changes in direction is often reduced to a minimum.

In terms of efficiency the circumnavigation should be initiated as early as possible as to avoid abrupt heading changes and thereto, to avoid unnecessary detours. With the limited knowledge of current weather due to the limited field of view, early initiation of a circumnavigation is usually not possible. Studies on a possible increase in efficiency due to increased weather information in the cockpit would be of interest. To this and some other issues the IMUK is devoted.


Development of weather avoidance model - DIVMET - for air traffic re-routing

Fig. 1: Concept of the algorithm DIVMET. The green line shows the given route of an aircraft. The grey sectors represent the aircraft’s radar field of view. If any weather object (blue) is recognized in this field of view and the aircraft will hit it or its safety margin/convex hull (red) on the given route, a decision is made and a rerouting via an heading change will happen in the model.

For the above reasons, the necessity to circumnavigate thunderstorms, a model – DIVMET (divert meteorology) – that simulates the avoidance behavior of aircraft in storm situations was developed at the Institute of Meteorology and Climatology. DIVMET is based on a proprietary path-finding algorithm MET2ROUTE that searches a path through a field of thunderstorms or other objects using evasive routes similar to types known from robotics. Additionally the principle of convex hull serves for trajectory calculation.
Weather is integrated in the simulation in form of two-dimensional polygons, so-called weather objects. These objects are created either generically or extracted from radar images. Their evolution with time and shift in the model is therefore dependent on the update rate of the underlying weather data. A projection into the future is not yet implemented. However, the integration of nowcasts is in progress and thus, the further development of the existing cells are taken into account in the model. In order to meet international standards, the weather objects are magnified by a safety distance. This can be changed to either obey the regulations (10-20 NM) or to represent observed behavior in which considerable smaller distances were kept. The resulting new objects are hereinafter referred to as risk areas. For routing around risk areas, concave regions are filled by using a convex hull.
Based on a planned route, an alternative route will be calculated when there is a conflict of the current route with a weather object or a risk area detected. With the goal of reproducing the actual avoidance behavior, a limited field of view was implemented. Analogous to the appearance of the radar field of view, the extent (the range and angle) can be specified in the model. This means that not right at the beginning of the simulation the whole weather situation is known on the planned route, but steadily expanded during the flight and a gradual adjustment of the route may be necessary. In contrast, the settings can also be selected so that at any time the complete weather information is known, whereby the advantage of broadened weather information in the cockpit can be examined on the efficiency of air traffic.
The decision in which direction a weather object is circumnavigated is based on the spatial extent of the risk area and possibly other objects left and right of the planned route. Is the larger extend located right of the route, the trajectory heads for the most outer point of the convex hull/ the risk area left of the route. The flight is then continued along the convex hull to a point, the tangent of which does intersect the original route in a predetermined or determinable point and the return to the planned route requires a change in direction of less than 30 °.
Output parameter of DIVMET is, in first place, the additional distance flown. Derived from this additional flight time, the volume of fuel consumed and fuel costs are issued.



Issues and objectives

  • Development of an understanding of the interaction of two systems (weather and air traffic) is based on a modeling approach
  • Simulation and representation of the current evasive behavior of the air transport users -> How well is the observed behavior reproduced?
  • Identification of the effect of extended weather hazard information in the cockpit on the efficiency of single trajectories
  • Identification of strategies for Air Traffic Management (ATM) to counter the stochastic nature of adverse weather phenomenons
  • Development of optimal strategies for route planning in unpredictable adverse weather conditions
  • Provide guidance for controllers and pilots to find a safe and efficient route in a field of adverse weather

Current applications

  • Simulation of arriving and departing traffic at Hong Kong International Airport and comparison with real data
  • Application of DIVMET to real planned data in the event of a Squall Line over Austria on July 17th, 2010 and comparison with real flown routes
  • Studies on the change in capacity and shift sector load in storm situations.

Not meteorologically related, but also relevant for air traffic and possible to simulate in the same way, are areas of volcanic ash that have to be avoided. Circumnavigations of air traffic around these areas can be model by DIVMET as well.



Current projects

  • Further development of DIVMET (Ludmila Sakiew, Manuela Sauer, Doctorate)
  • Simulation of the air traffic (DIVMET, NAVSIM) in case of an occurence of a squall line above Austria (Patrick Hupe, Master project)
  • First steps to account for aircraft-aircraft conflicts in DIVMET (Kezia Lange, Bachelor project)