Why do Soyuz landings seem to be overshot?

Question

Soyuz descent module re-entry flight is piloted (one of two modes AUS/computer or RUS/manual), using 8 thrusters, toward the Kazakhstan landing site. The ballistic re-entry occurs only when the other modes are inoperative (e.g. TMA-11) or in case of an urgent departure from the ISS.

With a controlled descent, the actual landing location should be equally distributed around the target point, but it seems most of the landings occur N or NE of it.

(source)

Typical flight path of the re-entering spacecraft:

(TMA-21, source)

As pointed out by TildalWave, the winds may cause the module to drift after the parachute is deployed. From the re-entry profile described at SpaceFlight-101:

At an altitude of about 9 Kilometers, the Pilot Chute opens and deploys the Drogue Chute that slows the vehicle from 240 meters per second to 90m/s. At an altitude of 7.5 Kilometers, the Main Chute is opened and slows the vehicle down to 6 meters per second. While flying under the main chute, the Soyuz transitions from a nearly horizontal flight to a vertical descent.

This gives about 8 kms of trajectory at low vertical speed (12 kts / 22 km/h) and high drag configuration where the module may drift. About the wind, as researched by Rikki-Tikki-Tavi, the prevailing wind blows to NE in Kazakhstan. The wind blows most of the time to NE and E, according to WeatherSpark for Kyzylorda Airport WX station:

At first sight, a correlation is possible. Buy why would the Russian Federal Space Agency ignore wind in the flight planning? Maybe this is their reasonning:

• The aircraft departure form the ISS is possible only when the wind is within given acceptable criteria which guarantee a drift cannot push the spacecraft outside a predefined safe area.
• A return schedule and flight path is computed without wind and published.
• Soyuz is deorbited according to this schedule.
• Some drift occurs due to the actual wind, the rescue team follows.

There are other possible explanations, e.g. this one:

• A delay occurs in the re-entry schedule.
• The predicted landing location is adjusted, but not shared with the public.
• The landing actually occurs at the adjusted location, but we don't know.

Neither of these scenario is really satisfactory. Is it possible to shed light on the way the re-entry path is managed, and whether the wind is taken into account?

Answer 1

Prevailing winds in Kazakhstan seem to be in a north-eastern direction. One factor may be that they are under-compensated in the targeting. Why, I cannot say. I think it is possible, though that the Russians are following a "never change a running system" approach here. Meaning they don't want to spend the money and manpower on fixing a problem that isn't more than a minor inconvenience to them.

Answer 2

They are not targeting what they think they are targeting.

I'm sure that there is a logical explanation (I don't think it's chance). However they apparently don't know the explanation. If they did, they would include it in their simulations and they would then be actually targeting what they think they're targeting, and the errors would be distributed more evenly around the target. I don't know the explanation either.

Though much smaller in number, I will note that our Mars landings tend to be long as well. That may be chance. If it isn't chance, then there is a systematic error there as well. Probably not the same systematic error as Soyuz. (All Mars entries except for MSL were ballistic. MSL was guided, like Soyuz.)

From there, all you can do is speculate. I'm sure that they do a very thorough job simulating all of the obvious things, so most of our speculations would be wrong. If we can think of it, chances are it's in their simulations with appropriate dispersions. Atmosphere density variations, wind, gravity field, aerodynamic coefficients of the entry vehicle, entry flight path angle, etc.