Dear editor,

While branched glycerol dialkyl glycerol tetraethers (brGDGTs) serve as critical terrestrial paleothermometers, their application in arid environments remains constrained by persistent confounding factors such as aridity, pH, and salinity. The manuscript by Dugerdil et al. makes a noteworthy contribution by systematically compiling brGDGTs datasets across Arid Central Asian and establishing a robust statistical framework. This work provides novel insights into brGDGTs proxy utilization under arid conditions and fills a critical geographic knowledge gap. However, several outstanding issues require resolution prior to publication:

Major concerns,

1. The research team has already published analyses of the 162 "new samples" in Paleoceanography and Paleoclimatology (https://doi.org/10.1029/2025PA005214). Continued characterization of these samples as "new" in the current manuscript (Abstract and Section 2) constitutes inappropriate claims.

1. This study presents a comprehensive statistical analysis of conventional brGDGT-based proxies. However, several prior investigations have consistently demonstrated that conventional temperature proxies such as MBT’_5ME exhibit reduced performance in arid regions compared to their efficacy in humid environments. To further advance the methodological contribution of this work, it is recommended to further evaluate recently developed temperature reconstruction approaches, such as Bayesian models and machine learning algorithms. Do these newly developed, global database–based methods also face limitations when applied in arid regions?

Minor comments,

Line 34: I would suggest removing the term "air" or replacing it with "ambient temperature", as brGDGTs primarily respond to soil or water temperature rather than air temperature.

Line 45: The CBT' should not be exclusively interpreted as reflecting cyclisation of branched tetraethers, as its definition inherently incorporates 6-methyl isomers. It would be scientifically more precise to either retain the original term "CBT" or adopt "CBT_5ME".

Line 55: The use of "thus" here appears strange.

Table 2 and Line 192: "CBT’_5ME" should be revised to "CBT_5ME".

Please see the attached file for my specific comments.