Lenghu on the Tibetan Plateau as an astronomical observing site

On Earth’s surface, there are only a handful of high-quality astronomical sites that meet the requirements for very large next-generation facilities. In the context of scientific opportunities in time-domain astronomy, a good site on the Tibetan Plateau will bridge the longitudinal gap between the known best sites1,2 (all in the Western Hemisphere). The Tibetan Plateau is the highest plateau on Earth, with an average elevation of over 4,000 metres, and thus potentially provides very good opportunities for astronomy and particle astrophysics3–5. Here we report the results of three years of monitoring of testing an area at a local summit on Saishiteng Mountain near Lenghu Town in Qinghai Province. The altitudes of the potential locations are between 4,200 and 4,500 metres. An area of over 100,000 square kilometres surrounding Lenghu Town has a lower altitude of below 3,000 metres, with an extremely arid climate and unusually clear local sky (day and night)6. Of the nights at the site, 70 per cent have clear, photometric conditions, with a median seeing of 0.75 arcseconds. The median night temperature variation is only 2.4 degrees Celsius, indicating very stable local surface air. The precipitable water vapour is lower than 2 millimetres for 55 per cent of the night.


2019.
A cumulative plot allows the simple extraction of important information, such as "what fraction of the time is the seeing better than 0.5 arcseconds?". Given the critical importance of the fraction of time when the seeing is superb, a cumulative plot is most informative. The plot should show for comparison the Dome A values, and those for Chile, Hawaii, etc.
Line 426, just a comment that the annual temperature variation at Lenghu is remarkably low, particularly the very low range during a day/night cycle.
Line 426, in fig 2, consider separating the data for 2019 and 2020 to make the comparison clearer.
Line 458, why show lines at 6 and 9 km? D. Appropriate use of statistics and treatment of uncertainties: All error bars should be defined in the corresponding figure legends; please comment if that's not the case. Please include in your report a specific comment on the appropriateness of any statistical tests, and the accuracy of the description of any error bars and probability values.
Yes, all fine, apart from the extraneous sig figs mentioned above. E. Conclusions: Do you find that the conclusions and data interpretation are robust, valid and reliable? Yes.
F. Suggested improvements: Please list additional experiments or data that could help strengthening the work in a revision. This paper introduces Lenghu area as a potential reservoir of sites for ground based astronomy in the eastern part of the Tibetan plateau. One of the summit has been thoroughly tested in the past two years, an impressive database of several key parameters has been collected and some results of the analysis are presented. The authors aim at comparing the astroclimate properties of this site to those of a few major reference ground based observatories in operation. They show convincingly that it is in many aspects of comparable quality for future scientific projects. The text is clear and the conclusions are well documented, I consider nevertheless the paper is too modest and should be expanded in some domains to reach the level of expectancy of Nature readers, suggestions are detailed in what follows.

--Introduction
Many readers are not aware of astronomical site selection activities in China, putting the present survey in a global perspective would be helpful: the project should mention in addition to Ali, the other potential sites in the western Tibetan plateau which were part of the site search for the Chinese 12m project. It is obvious, looking at the attached map, that Lenghu had a unique geographical position, filling a gap in the area.
--Water Vapor Ground based observations in the infrared suffer of the background noise created by the radiation of the telescope structure and components, in particular on "warm" sites like in the coastal Atacama. Hence the very low water vapor period can only be used if the telescope is cooled, either actively which is unrealistic for ELTs, or passively if the site is cold enough. It appears that Lenghu has an enormous advantage in this respect, being the coldest of the Tibetan sites in winter (Ali median is -10C, the 2 others -5C) precisely when the PWV is at the lowest. Moreover, with global warming predictions of about .5C per decade (attach.2), some of the warmer sites could well reach positive winter temperature within this century.
--Seeing The authors have collected an impressive amount of continuous DIMM (please correct page 3 line 82 DIMM = Differential Image Motion Monitor) data which should allow to extract much more than the median 1mn seeing: the median seeing expected to be achieved as a function of the exposure time is also relevant. As it is clear that the Lenghu seeing is variable with a long tail in the distribution, it is important to know if the spikes occur as groups or rather aleatory. An example of such an analysis is given by R. Racine in Mauna Kea in "Temporal fluctuations of atmospheric seeing" 1996, PASP 108, 372-374. Secondly, it is important to be able to estimate which fraction of the seeing is close to the ground, hence removable on wide field with GLAO technique. The seeing rose, polar plot of the seeing records versus simultaneous incoming wind direction, can help. Should the bad seeing be always linked to the same wind direction, it is clearly caused by local features and thus located close to the surface.
Congratulation to the Lenghu monitoring team for this achievement Author Rebuttals to Initial Comments:

Dear Reviewers,
We greatly appreciate your efforts in the previous versions of the manuscript. We have made point-to-point responses to all the comments/suggestions raised in your review reports and made the corresponding revisions in the context. All the replies in this document are colored by blue and the revisions/changes in the revised manuscript are marked in red.

Sincerely yours,
The authors.

General Response to Referees' comments:
To facilitate our response, a map provided by Referee #2 (we appreciate that very much!) is attached here to illustrate the three candidate sites for the Chinese 12m LOT project together with Lenghu. This map shows where exactly the sites are located, and the relative accessibilities among them in the context of logistics and support for future developments.
The blue lines only connect the sites, to direct eyes. They are all state roads (mostly paved), except the Ngari-Tashkurgen part, which is a gravel road, and sometimes just truck tire prints in wilderness. Please also refer to the map in the extend data part for the transportation manners to reach Lenghu.

Response:
Thank you for paying attention to site survey activities in West-China.
More words are needed to explain your concern, but first of all, thank you for pointing out this.
Since 2000, site surveys (restarted many times for complicated issues, one of them has been funding) have been carried out. The tested sites are: Ngari (Ali), Muztagh Ata, Daocheng (indicated on the map above), and other spots also on the plateau tested only for very short periods and then abandoned. These actions were all before 12m LOT proposal (2015). Then, the three sites were officially included in the grand plan of LOT in 2016. The site testing at Lenghu is a separate task, independent of LOT site project. The first author of this work (LD) is one of the 3 "external observers" appointed by LOT committee. In 2020 the site team published 10 papers (RAA Vol20 No6, the overview paper cited in the current work) as a mini volume. I was then invited to review the manuscripts. For this reason, there is no common names between those papers and the current one.
[Redacted]. One can never imagine (me too) that the Daocheng site was selected (by vote) for 12m LOT in the latest proposal (in 2019, when Lenghu is still being tested). As in the overview paper (ref24) of the mini volume in RAA, Daocheng is the worst one among the 3 sites even in common sense (unusable summer, and very humid winter).
Ref7 (now ref.8) is a Science journal report about the early site testing in Ngari region. From the report, the Japanese participants raised a few issues on light pollution and strong wind at the site. As of today, the local city brightened up a lot, and the subsequence testing work did not go very smoothly. Later in 2018, an astro-particle team started a project for Primodial Gravitational Wave (PGW) project (new ref.7, also a Science journal report by Normile in 2019), claimed a super good site. However, that is not really "Astronomical Observatory". As it turned out later, that both city light of Shiquanhe and strong winds are troubles. Many telescope projects originally planned for Ngari station were moved out from that site, some of them (telescopes of Nanjing University time-domain project, for instance) are recently relocated to Lenghu site.
Lenghu, as a future site, was never into the 12m LOT scope in the previous site selection campaign, the main reason was that Lenghu site monitoring started slight later (end of 2017) than the 3 sites. However, as of the end of 2020, Lenghu attracted a lot of attention from the decision-making level, mostly due to its extraordinary results in comparison to the 3 sites of LOT, and in addition the excellent accessibility. In the latest development (after this paper's submission), A Magellan type 6.5m telescope is being considered to come to Lenghu; LAMOST (currently in Xinglong) will be upgraded to 8m aperture/15k fibers, and then will be relocated also to Lenghu. Although 12m LOT proposal is still waiting for final funding, but once it will be decided, I believe it will also come to Lenghu (already being discussed at NAOC and CAS, the institution responsible for building and operating LOT). This makes this work even more timely.
A table comparing Lenghu and the three sites in terms of seeing and observing time is now added to the extended data part, based on a processed data set provided by LOT site paper.
Seeing measurements for all the 4 sites were taken using the same instruments and software.
Observing time estimation are made on the same method of LOT in this comparison.

Revision in the manuscript:
(Lines 151-154) In terms of total seeing, Lenghu is comparable with best established sites (in Chile, Hawaii, and the Canary Islands), and is obviously the best one on the Tibetan Plateau (see Methods, Extended Data Table 2, and Extended Data Figure 4). On the surface of the earth, the best conditions in both seeing and observing duty cycle for time-domain is clearly on Antarctica 5,6 .

Comparison of Tibetan sites: key parameters
In addition to Lenghu, 3 other sites on different locations on the Tibetan Plateau were also tested for different purposes since 2000. An intensive site testing program was carried out for Chinese 12 m telescope recently at these 3 sites, namely Ngari, Muztagh Ata and Daocheng, and are concluded in the overview paper 24 . It turns out that Lenghu has the best observing condition for optical/infrared. The direct comparisons of the key parameters of AOT and seeing are shown in Extended Data Table 2 and Extended Data Fig. 4, respectively. It is noted that, since the seeing data at Ngari, Muztagh Ata and Daocheng are all truncated at 3.0 arcseconds, the seeing data at Lenghu is also truncated at 3.0 arcseconds in Extended Data The data and methodology of the paper seem perfectly fine. There are a dozen or so typos and other editorial changes that I haven't bothered to mention since the Nature editors will find them quickly. I list below some comments, with line numbers.

Response:
Thanks. We already spotted the typos, all corrected.
Line 30, the article compares Lenghu with the "best sites known for astronomy across the world", but does not mention Antarctica, which is well-established now as containing considerably better sites for most types of astronomy. In particular, China itself has Kunlun Station at Dome A, which is at the best location within Antarctica, and the subject of a recent Nature article (Ma, et al. 2019, Nature 583, 771-774). This paper is in the reference list, but not discussed in any detail. The authors should acknowledge more fully the existence of

Dome A at least, and the superior site properties there. The reason that Dome A is not a suitable choice for China's large telescope is simply that the logistical challenges of building
it there are too great, without a pathfinder smaller-scale telescope such as the proposed KDUST telescope.

Response:
Thank you for mentioning this, we have good connection with Zhaohui during this work. we highly valued the Dome A paper. We regarded Dome A as a quasi-space program, so that we did not even try to match the results. But as you suggested, a short description is added.

Revision in the manuscript:
(Lines 151-154) In terms of total seeing, Lenghu is comparable with best established sites (in Chile, Hawaii, and the Canary Islands), and is obviously the best one on the Tibetan Plateau (see Methods, Extended Data Table 2, and Extended Data Figure 4). On the surface of the earth, the best conditions in both seeing and observing duty cycle for time-domain is clearly on Antarctica 5,6 .
Line 46, yes, the unique geographic position of Lenghu is certainly a big advantage. It would be worth mentioning that Dome A has even better time-coverage than Lenghu, due to its location close to the South Pole, although there is an obvious trade-off between the length of time that you can observe a given position on the sky, and the amount of sky coverage that you have.

Response:
Thank you. Regarding time-domain science in connection with Dome A and Dome C, some wording is included (Lines 151-154).

Response:
Thank the reviewer for the comment. We estimated the area with altitude below 3,000 m around Lenghu Town using the tomographic data presented in Extended Data Figure 1. The area is over 100 thousand square kilometers.

Revision in the manuscript:
(Lines 54-56) The huge area (over 100,000 square kilometers) surrounding Lenghu Town has a relatively low altitude below 3,000 m, the climate is extremely arid, like the Atacama Desert, and the local sky (day and night) has long been known to be unusually clear 12 .
Line 103, the SQM does not accurately measure light from all angles above the horizon as the article states. It is quite peaked towards the zenith, and it has less that 20% of its peak responsivity at zenith distances greater than 60°. The manufacturer says that the effective solid angle is about 1.53 rads, which is only 24% of a hemisphere.

Response:
Yes, the model SQM-LE deployed at Lenghu follows the specifications of the manufacturer.
It detects light from lower altitudes (zenith angle > 60) at very low efficiency. We changed wording in the text.

Revision in the manuscript:
(Lines 104-106) …accurately measures the integrated light of the entire visible sky, with the sensitivity optimized toward the zenith and quickly drop-off to less than 20% once the zenith angle is greater than 60°.

Response:
Thank you for the comment. We have specified this in the manuscript. In our plan, a MASS (being built by our colleagues from NIAOT) was expected to operate no later than middle of 2020. Unfortunately, the pandemic slowed down quite a few things in our project including this one.

Revision in the manuscript:
(Lines 131-134) Measurements by DIMM are nowadays widely used and have become a standard assessment for integrated atmospheric optical turbulence 5,16,17 . Multi-aperture scintillation sensor (MASS) can provide further information on how the total seeing is composed of from different layers in the atmosphere above the site 6 .

Line 132 should specify the exact model of DIMM used; ref 24 talks about three different models, and ref 25 describes a range of instruments, so it is not possible to determine
unambiguously the model used.

Response:
Thank you. Ref. 24 applied 3 different types, one of them is referred as "French DIMM". The other two are either manufactured by NIAOT or home-made by the Ngari team. Intercalibrations were made between all the systems before and during the campaign. The "French DIMMs" used by LOT team and by us are all from Alcor-System which comes with the same optical/mechanical design and software, as in ref25. It is the only full-system model offered by the company.

Revision in the manuscript:
(Lines 134-136) The same make of DIMM (the one provided by Alcor-System) applied in this work is also used for many different site selection campaigns 24,25 . Measurements are calculated for average wavelength 550 nm and corrected to the zenith (airmass unity).

Response:
The DIMM at Lenghu site operates at 550 nm. We have double checked with the manufacturer Alcor-System. The wavelength of 500 nm was adopted in ref. 24, I believe that is not correct.
Lines 155, 162, 163, 260 the observatory scores should be given to 2 sig figs. The 3rd digit is not justified (i.e., annual variations, and some details of the comparing the data from the sites, will certainly cause it to fluctuate).

Response:
You are right, thanks. All the values are given to 2 sig figs.
Line 305, the data are supposed to be available to the public here: https://lenghu.chinavo.org/index.html but this website gives a security warning, and when bypassed, it then gives a "404 Error -Page Not Found". This should be fixed prior to publication.

Response:
Thank you for pointing out this. For security reasons, our web administrator shutdown all ports but the classical http ports: 80 and 8080. Therefore, one should always use http instead of https. We have double-checked the data link by inviting several domestic and foreign persons to click the link. All of them can access the website normally.
In case the problem will be persistent, we can use other way to make our data publicly available.
Line 310, the Japanese data site requires a username/password; perhaps indicate how this can be obtained.

Response:
Thank you for pointing out this. According to data policy of JAXA, we cannot distribute the data directly. But one can easily register a username and password to freely download the data for academic use. We have added a sentence in the Data Availability to point out this.

Response:
We greatly appreciate your suggestion. Fig. 1 has been revised to display the histogram and the cumulative curve. The fraction of time that the seeing is better than 0.51, 0.61, 1.03, and

Response:
A good suggestion, thanks. The annual temperature variation at Lenghu is remarkably low, in favor of astronomical observations. We have revised Fig. 2  Line 458, why show lines at 6 and 9 km?

Response:
At an altitude of 6-9 km, the turbulence profile shows a clear difference in August and November, which suggests possible seasonal changes. We added three sentences in the paragraph of 'Turbulence Profiles'.

Revisions in the manuscript:
(Lines 303-307) Above 11 km, C 2 N decreases monotonously with no seasonal pattern. C 2 N is around 10 -17.5 and 10 -17 between 4 km to 11 km. On November 16, the two turbulence profiles show a similar trend, but the turbulence strength at night (red profile) is lower than that in the morning (grey profile). At an altitude of 6-9 km, the turbulence profile shows a clear difference in August and November, which suggests possible seasonal changes. Yes, all fine, apart from the extraneous sig figs mentioned above.

Response：
Thank you for this comment. We have revised the sig figs in the text. For all figures involve statistics of data, either error bars or quantile or percentile are now given. Thank you very much, Michael, for the valuable comments.

Referee #2 (Remarks to the Author):
This paper introduces Lenghu area as a potential reservoir of sites for ground based astronomy in the eastern part of the Tibetan plateau. One of the summit has been thoroughly tested in the past two years, an impressive database of several key parameters has been

Response:
Thank the reviewer for the comment. Text adjusted. We thank you for providing this map, which is also used to show to the editors and the other referee.
We have added text regarding LOT site surveys and provided quantitative comparison (figure and table) in the Extended Data.

Revisions in the manuscript:
(Lines 151-154) In terms of total seeing, Lenghu is comparable with best established sites (in Chile, Hawaii, and the Canary Islands), and is obviously the best one on the Tibetan Plateau (see Methods, Extended Data Table 2, and Extended Data Figure 4). On the surface of the earth, the best conditions in both seeing and observing duty cycle for time-domain is clearly on Antarctica 5,6 .

Comparison of Tibetan sites: key parameters
In addition to Lenghu, 3 other sites on different locations on the Tibetan Plateau were also tested for different purposes since 2000. An intensive site testing program was carried out for Chinese 12 m telescope recently at these 3 sites, namely Ngari, Muztagh Ata and Daocheng, and are concluded in the overview paper 24 . It turns out that Lenghu has the best observing condition for optical/infrared. The direct comparisons of the key parameters of AOT and seeing are shown in Extended Data Table 2 and Extended Data Fig. 4, respectively. It is noted that, since the seeing data at Ngari, Muztagh Ata and Daocheng are all truncated at 3.0 arcseconds, the seeing data at Lenghu is also truncated at 3.0 arcseconds in Extended Data In Extended Data Table 2, the AOT is calculated using the LOT method based on all sky camera images. They divide the total visible sky by two circles with zenith angle 44.7° and 65°, namely the inner and the outer circle. When there is no cloud in the outer circle, it is defined as 'clear' (otherwise photometric in ref. 24); and it is 'outer' if only the inner circle is