This manuscript presents an important update and improvement based on a widely recognized dataset. I commend the authors for their excellent work. Building on the C-LSAT 2.0 dataset, the authors have integrated nearly 3,000 additional global observational stations to develop the C-LSAT 2.1, along with high-resolution (0.5°×0.5°) LSAT and DTR datasets. The new version significantly enhances data coverage, refines processing techniques, and provides valuable high-resolution products. The manuscript is comprehensive, logically structured, and the conclusions are sound. It fits well within the scope of Earth System Science Data (ESSD). However, the following minor revisions are recommended prior to publication:

1. Please supplement the temporal coverage of observation stations at both annual and monthly scales.

1. In section 4.1.2, the authors used the MAE, RMSE to evaluate the different datasets. However, sometimes, MAE and RMSE only provide one aspect performances of different datasets. DISO as a comprehensive performance evaluating index can illustrate the overall performance for all the datasets. Therefore, the DISO should be added in this study.

1. If feasible, consider adding additional validation in representative regions, particularly in high-altitude or complex terrain areas.

1. All variables in equations should be consistently italicized.

1. Insert appropriate spaces between numbers, symbols, and units (e.g., in line 230).

1. Add a space before parentheses in Table 1 for clarity.

1. Please explain the observed increases in MAE and RMSE since 1990 as shown in Figure 7.

1. The abnormal warming pattern in northern North America depicted in Figure 12(d) requires verification.

1. The colorbars in Figures 11 and 17 hinder regional comparisons; consider revising them for clarity.

1. There are citation errors that need correction (e.g., in line 776).

Overall, with these minor adjustments, the manuscript will make a valuable contribution to the climate data community.

General comments

The manuscript describes the update of a valuable dataset of Land Surface Temperature through the inclusion of a significant number of additional stations, and the creation using the TSP method of high-resolution datasets of land surface temperature and diurnal temperature range. The manuscript is clear and scientifically sound. I have two main concerns that should be addressed before publication:

- it’s not clear whether the trend uncertainties (e.g Line 230) have been computed taking the serial correlation of the time series into account, as the time series (e.g. Figure 2) seem to display a significant autocorrelation/serial dependence, which biases downwards the uncertainties if not accounted for. This is something that should be clarified.

- the quality of the figures should be improved, to make them easier to understand, particularly in case of colour vision deficiencies.

Minor comments and suggestions are described below.

Specific comments:

Line 138: maybe give some more details on the filtering procedure

Figure 1: The colours in Figure 1 are very similar for some of the datasets and very difficult to distinguish – particularly for I suggest using more contrasting colours and different types of line (e.g. dashed, dotted) to make the figure clearer.

Line 166: maybe use : before “Any anomaly…” to make more obvious that the sentence is referring to the quality control process.

The same notation could be used to denote standard deviation, STD is used in line 177, and sigma in line 184

Table 1: the caption and text should be improved, to make clear what is exactly shown in the table – maybe instead of results of QC something like number of data values excluded during the QC procedure? The “unit:station month”could also be made clearer.

Line 221: enhance (instead of “ which significantly enhancing”)

Figure 2: same as for Figure 1, it’s hard to distinguish the different colours / lines

Line 230: please indicate if autocorrelation was taken into account in the estimation of trend uncertainties

Table: Ele was not defined for elevation in the text

Figures 6, 7: the figure would be clearer with a slightly larger size of the text in the axis and legend; using different line styles (dashed, etc…) would make the figure easier to perceive in case of colour vision deficiencies.

Figure 9: the colour scale should be improved, ensuring it is centred on zero and has improved perceptual properties

Figure 10: same as previous figures, using different line styles (dashed, etc…) would make the figure easier to perceive in case of colour vision deficiencies

Figures 11, 12: ideally, the diverging colormap scales should be symmetrical

Table 5: is the * denoting statistical significance? Was the uncertainty computed assuming any form of linear dependence or just white (uncorrelated) errors??

Figure 15: the colour scale should be improved – differences in colour between 20-25 ºC and 25-30ºC are not distinguishable