Seeing effects in the wake of a building
(summary of previous work 1990-1991)

Scope of the study:
Impact of a new telescope building on the observing quality at the Observatory
Search for the optimal location of the new unit

By: Marc Sarazin, ESO Garching, May 2000

1. Initial Assumption:

Various flow patterns contribute to the wake of a building: heat input from the ground by recirculation inside the cavity zone is the main source of thermal turbulence in the wake.

2. Seeing in the wake of a thermally neutral building: simulations

(Source: VLT Report VLT/TRE/RIS/17400/0001, A. de Baas, RISOE Denmark, March 1991)

The temperature structure function (Ct2, units 0.001 K^(2) m^(-2/3)) in a vertical plane aligned with the wind. A H=30m high building (type VLT-UT) is at y=0. Four regions can be identified: upwind, top, wake and recovered region. Although the inlet flow is recovered at a very large downwind distance (X=20H), only at the top of the building is a significant increase of Ct2 noticeable due to flow lift-off (this effect is reduced in an open structure which has a smaller apparent height).

The temperature structure function in a horizontal plane behind a row of four 30m high buildings (type VLT-UTs) at y=-50,-14,18,51 perpendicular to the wind. The inlet Ct2 is 6.4, 4.3, 2.8 at 5, 10 and 25m height respectively. The simulation shows a reduction of the Ct2 in the wake in the first 20m above ground due to mechanical turbulent mixing.
The ground thermal input is not included in the model.

3. Seeing in the wake of a thermally neutral building: measurements

Seeing measurements at a distance (x=2H) of the NTT building at La Silla were compared to the atmospheric seeing using two DIMM monitors simultaneously . Up to 0.5" additional seeing was detected at ground level in low wind conditions, because of heat exchanges in the recirculation zone. For wind velocity above 4m/s, the contribution of the wake becomes negligible or even favourable.

Excess seeing (arcsec) versus wind speed (m/s) when observing through the wake at a distance x=2H of the center of the building (Source: VLT Report 62, Appendix A, Nov. 1990, Ed. M. Sarazin).

4. Seeing in the wake of a thermally active building: measurements

(Source: Trimestrial Meteorological Report 72bis, M. Sarazin, April 1991)

Ct2 measurements at a distance (x=2H) downwind of the NTT building (H=20m) at La Silla were compared to Ct2 measurements at a distance (x=2H') downwind of 3 nearby non insulated water tanks (H'=H/4) at various heights above ground (2H', 4H', 5H')). The Ct2 in the wake of the 4 times smaller water tanks is found to be 10 times larger than in the wake of the NTT building. The water tanks never reach thermal equilibrium with the ambient air.

5. Conclusions

Assuming that a telescope building is by design thermally neutral (no heat exhaust, light structures, flushed enclosure), the far wake of the building, extending horizontally up to 20 times the building height, is not damaging to transverse optical propagation. The only potential source of optical turbulence is between the ground and the building top, in the recirculation area of the near wake of the building. As it is the case in mirror seeing, the ground to air thermal advection is proportional to their temperature difference and decreases for larger wind speeds.
For an absolute protection of the observatory environment, the recirculation area downwind of a telescope should ideally be covered by an insulating layer of short external thermal time constant. The recirculation area extends from immediately behind the building to a distance of min 1.5, max 2.5 times the largest dimension of the building, depending on the building shape and roughness.