Passer par le Nord X Meltemus: Routing through the ice

Preparing for an Arctic circumnavigation by sail, without motorized assistance, imposes a particular requirement: being able to navigate through areas subject to the presence, sometimes massive, of sea ice, without taking unnecessary risks. However, it is impossible to dynamically integrate ice density into the routing calculations performed by routing software such as Adrena or qtVlm. 

The problem: an "ice" constraint that is impossible to exploit

For my part, I use QtVlm, a free navigation software developed by Meltemus (Philippe Lelong), which allows weather routing, polar curve management, gribs, routes, and numerous sensors. While preparing my routes to simulate Arctic circumnavigation routes, I noticed that although qtVlm could display ice density data from GRIB files (such as those published by AMSR2 or NOAA), it was not possible to integrate them directly as constraints in the routing. I was left with two inadequate alternatives:

• create fixed barriers (using shapefiles) that do not take into account changes in ice over time;
• or ignore ice in the routing calculation, which is obviously unthinkable.

Proposed solution: speed dependent on ice density

I proposed a simple and workable solution to Philippe Lelong, lead developer of qtVlm: introduce a formula to slow down speed based on ice density, with a maximum density beyond which speed drops to zero.

Proposed formula:

Vb=Vm×(1−Dg/Dgmax)Vb=Vm×(1−Dg/Dgmax)

• Vb: effective speed in ice
• Vm: speed in open water (extracted from the polar curve)
• DG: ice density at time T
• Dgmax: maximum passable density defined by the user

This formula has two advantages:

1. Simplicity of implementation: only one user parameter to define.
2. Algorithmic elegance: the routing engine naturally "avoids" areas where speed drops to zero, without artificial constraints.

Implementation: a quick and effective response

Philippe responded positively. In less than a week, a beta version of qtVlm incorporating this new logic was released. The concrete benefits:

• New "Ice Density" field visible in the speed table and route histograms.
• Speed reduction coefficient visible when flying over the STW field.
• Gradual reduction in speed depending on ice density.

The system works: during testing, routes avoid high-density areas without manual intervention. Readability is still hampered by the Mercator projection in high latitudes, but this could evolve towards a Lambert projection in the future.

Feedback and contributions

During my testing, I reported several minor bugs (roads disappearing when closing, erratic behavior on certain long routes) and suggested aesthetic improvements (recalibrating the color gradient for windows). All of these contributions were incorporated with exemplary responsiveness.

‍