I have read the manuscript « Tropical temperature evolution across two glacial cycles derived from speleothem fluid inclusion microthermometry » with great interest ; it was pleasant to read and well organised.

The authors present a reconstruction of tropical land temperatures derived from nucleation-assisted  fluid inclusion microthermometry on a stalagmite (WR5_B) from northern Borneo. The record spans MIS 12 to MIS 9, covering two glacial terminations and one glacial inception (127 kyr in total). The authors analysed an extensive set of 2090 fluid inclusions in 49 different growth bands from that sample and an additional 146 fluid inclusions in 3 growth band from a late Holocene sample for comparaison. Previous work showed Borneo temperatures tracked CO₂ and Antarctic temperatures during the last termination. This study tests if that pattern holds for earlier terminations and glacial cycles.

The manuscript represent a significant contribution to the knowledge of past climate in the tropics over a long, less often studied timescale (from 460 to 333 ka) and the amount and replication of fluid inclusion (FI) microthermometry measurements give confidence in the results. The applied method of FI microthermometry on stalagmites is still regarded as novel in the field of speleothem science although it has been around for some time because it is not commonly use. To date, when it is applicable, it is the most accurate palaeothermometer available on speleothems.

The resulting temperature record is of high quality and precision and the discussion of the results is appropriate and well put into context. The only impair is the age model of stalagmite WR5_B resulting in uncertainty on the order of ±12 kyr. Although this can be regarded as large for speleothem chronology, it is comparable to other archives available for the time period (i.e. marine sediment) and taken into account into the discussion of climatic events. This new record shows that land temperatures closely followed Antarctic temperature changes during Terminations V and IV, consistent with previous findings from the last deglaciation. Deglacial warming began with Heinrich stadials, marked by drying in Borneo and Northern Hemisphere cooling, confirming a decoupling between temperature and hydroclimate in the region. Finally the authors have shown that a strong linear relationship between Antarctic temperature and Borneo records suggests a consistent polar amplification factor for the period of the LGM to the Holocene and  MIS 12 to MIS 9.

I would therefore recommend the acceptance of this manuscript in Climate of the Past after a few minor corrections.

Minor comments :

Line 98 – 99 « Although the two temperature records do not cover the same time interval it is safe to assume that present-day cave temperature in Whiterock Cave closely reflects (multi) annual mean outside air temperature without seasonal bias. »

Here the argument could benefit or be supported by a check of the climatological reanalysis data for the same time interval.

Line 147 we additionally took into account 

And « major hiatus (red dashed) » : is it the growth hiatus shown in Fig A1 ? If yes it would be good to refer to the figure ;

Figure 4 : is this inclusion from the studied stalagmite ? could you mention from which stalagmite it comes from ?

Line 320 : you mention the lapse rate uncertainty, could you give number ?

Line 349 : « we employed our temperature record to calculate the oxygen isotopic composition of the calcite supplying drip water d18Odw from which the stalagmite formed » I think this sentence need to be re-written ; isn’t the drip water supplying speleothem formation and not the calcite ?

Line 351 : Why did you use the empirical speleothem calibration for the calcite-water oxygen isotope fractionation proposed by Tremaine et al. (2011) ; and not other equations like Kim and O’neill (1997) ? I would be interesting to see the difference between the most commonly used equation in the litterature. Maybe this choice is based on modern monitoring data, if so, it would be worth mentionning it

Line 371 : « the previous interpretation of relative drying » not sure what « previous » refers to here ? Maybe a reference is needed

Line 392 : That is correct but it would be good to give estimates of uncertaities on both record

Line 459 : Give altitude difference in paranthesis

Figure 8 : You discuss above (between line 459 and 469) that you corrected for the difference in cave altitude between the Secret Chamber cave (Loland 2022) and Whiterock Cave (this study) so why not plotting a corrected version to make comparison easier for the reader ?

Why termination T-IV is not compared with T-I and T-V here ? I understand there is a hiatus during part of the glacial-interglacial transition but this hiatus also holds information ; is the timing of the hiatus comparable to an HS event-like ? A hiatus would fit with your argument of relative drying during Heinrich Stadials

Figure 9 : How did you select individual datapoint ? Is it based on each individual FI measured datapoint from your record and then the corresponding T° from the Antarctic record ? Or is at a point at a regular time interval from your interpolated record ? Could you clarify in the figure caption ?

Figure A4 : There are two different plot with Sample Pos. 29 (page 28 and 29)

Stalagmite WR5_B has previously been studied in a comparison study of different speleothem paleo thermometers (Meckler et al., 2015). I saw that the FI stable isotopes are plotted in figure 5 ; but what about the previous FI microthermometry data ? How does this newly produced data compare to the one in the study of Meckler et al. (2015) ? It could be briefly mentionned in the result section if uncertainty improvement have been made since that study, or shown in a supplementary figure in the appendix (showing the 2015 and 2025 data for the time period they overlap).

General comment : The notation of all temperature have to be made homogeneous (either space or no space between number and ±)

Krüger et al. present a spectacular record of temperature reconstructions from Borneo spanning two glacial-interglacial cycles. The manuscript improves our understanding of tropical paleoclimate in a region where absolute temperature records are scarce. The authors present their findings with clarity and appropriate context. This paper will be of broad interest to the paleoclimate community and therefore merits publication. However, there are a few minor issues to address:

• Typo on Fig. A2: Cheng et al. 2016
• The authors may consider moving Fig. A2 to section 2.2, since many of the millennial-scale climate interpretations hinge on the detailed wiggle matching to Cheng et al. 2016
• Section 3.2 left me wondering how the globally averaged sea level reconstructions from Bintanja et al. compare to regional sea level reconstructions off the coast of Vietnam, New Guinea, etc. Do small differences in regional vs global sea level at, for example, the LGM yield significant differences in temperature corrections? If so, how can these differences be appropriately accounted for in the T_corr uncertainties deeper in time?
• Although not my specialty, the term Polar Amplification is typically used to describe the enhanced polar temperature response relative to the tropics. This study presents data from the Western Pacific Warm Pool alone, which may not be representative of the broader tropical average. Clarifying this distinction will help avoid overgeneralization. The authors may also consider using the term Antarctic Amplification as opposed to polar.