Thirty Meter Telescope Site Testing VI: Turbulence Profiles

The results on the vertical distribution of optical turbulence above the five mountains which were investigated by the site testing for the Thirty Meter Telescope (TMT) are reported. On San Pedro Martir in Mexico, the 13 North site on Mauna Kea and three mountains in northern Chile Cerro Tolar, Cerro Armazones and Cerro Tolonchar, MASS-DIMM turbulence profilers have been operated over at least two years. Acoustic turbulence profilers - SODARs - were also operated at these sites. The obtained turbulence profiles indicate that at all sites the lowest 200m are the main source of the total seeing observed, with the Chilean sites showing a weaker ground layer than the other two sites. The two northern hemisphere sites have weaker turbulence at altitudes above 500m, with 13N showing the weakest 16km turbulence, being responsible for the large isoplanatic angle at this site. The influence of the jetstream and wind speeds close to the ground on the clear sky turbulence strength throughout the atmosphere are discussed, as well as seasonal and nocturnal variations. This is the sixth article in a series discussing the TMT site testing project.

💡 Research Summary

The paper presents a comprehensive analysis of atmospheric optical turbulence profiles measured at the five candidate sites for the Thirty Meter Telescope (TMT): San Pedro Martir (Mexico), the 13 N site on Mauna Kea (Hawaii), and three northern Chilean sites (Cerro Tolar, Cerro Armazones, Cerro Tolonchar). Over a minimum of two years, each site was equipped with a combined MASS‑DIMM instrument and, where possible, acoustic turbulence profilers (SODAR). MASS provides integrated Cn² values at six discrete altitudes (0.5, 1, 2, 4, 8, 16 km), while DIMM yields the total seeing from the telescope level to the top of the atmosphere. By comparing simultaneous MASS and DIMM measurements, the authors derive the ground‑layer (GL) seeing (the turbulence in the lowest ~200 m). SODAR data, covering 10–800 m with 5–20 m resolution, are used to validate and extend the low‑altitude portion of the profiles.

Key findings are: (1) The GL dominates the total seeing at all sites, contributing roughly half or more of the integrated turbulence. The Chilean sites exhibit a 30–50 % weaker GL than the northern‑hemisphere sites (13 N and SPM). (2) Above ~500 m, the northern sites have generally weaker turbulence, but 13 N shows the lowest turbulence at the highest MASS layer (16 km), which translates into the largest isoplanatic angle (θ₀) among all sites. Because θ₀ weights turbulence by h⁵⁄³, high‑altitude layers have a disproportionate effect. (3) The jet stream strongly influences turbulence in the 8–16 km range, with seasonal peaks in winter (both hemispheres) and reduced activity in summer. (4) Ground‑level wind speed correlates positively with GL strength; stronger winds increase low‑altitude turbulence. (5) Seasonal and nocturnal variations are evident: winter months show elevated high‑altitude turbulence, while a modest rise in GL turbulence occurs just after sunset and before sunrise, likely linked to surface temperature changes.

The authors discuss two approaches to constructing “representative” turbulence profiles. The first uses median and mean values of each MASS layer over the entire dataset, providing a statistical overview but ignoring inter‑layer correlations. The second selects subsets of data corresponding to specific percentiles of integrated parameters (seeing, θ₀, coherence time τ₀) and computes the associated layer statistics, yielding profiles that are more relevant for adaptive‑optics system design. The contrast between these methods highlights how different weighting schemes (equal weighting for seeing versus h⁵⁄³ weighting for θ₀) affect the derived profiles.

Cross‑validation between MASS‑DIMM and SODAR shows agreement within 10 % for the overlapping altitude range, confirming the reliability of the acoustic measurements despite occasional noise (notably at SPM due to surrounding vegetation). The combined dataset reveals a crossover in turbulence strength below 200 m: while the Chilean sites have weaker GL, they possess stronger turbulence aloft compared to the northern sites, explaining the observed differences in total seeing and isoplanatic angle.

Overall, the study provides high‑resolution, long‑term turbulence statistics essential for the design and optimization of adaptive‑optics systems for the TMT and other extremely large telescopes. By quantifying the vertical distribution of turbulence, its seasonal behavior, and its dependence on wind, the authors supply the necessary inputs for performance modeling, site selection, and operational planning of next‑generation astronomical facilities.