TRIMS LST: a daily 1&thinsp;km all-weather land surface temperature dataset for China's landmass and surrounding areas (2000–2022)

Tang, Wenbin; Zhou, Ji; Ma, Jin; Wang, Ziwei; Ding, Lirong; Zhang, Xiaodong; Zhang, Xu

doi:https://doi.org/10.5194/essd-16-387-2024

Articles | Volume 16, issue 1

https://doi.org/10.5194/essd-16-387-2024

© Author(s) 2024. This work is distributed under
the Creative Commons Attribution 4.0 License.

https://doi.org/10.5194/essd-16-387-2024

© Author(s) 2024. This work is distributed under
the Creative Commons Attribution 4.0 License.

Articles | Volume 16, issue 1

Data description paper

|

17 Jan 2024

Data description paper |

| 17 Jan 2024

TRIMS LST: a daily 1 km all-weather land surface temperature dataset for China's landmass and surrounding areas (2000–2022)

Wenbin Tang, Ji Zhou, Jin Ma, Ziwei Wang, Lirong Ding, Xiaodong Zhang, and Xu Zhang

Download

Final revised paper (published on 17 Jan 2024)
Preprint (discussion started on 22 Mar 2023)

Interactive discussion

Status: closed

RC1:
'Comment on essd-2023-27', Anonymous Referee #1, 24 Apr 2023

Dear Editor,
The manuscript reported a daily 1-km all-weather LST dataset for the Chinese landmass and surrounding areas from 2000 to 2021 with the special features of the longest time coverage, high image quality, and good accuracy. The study is interesting and fits within the scope of “Earth System Science Data”.
The main concept of the paper (the necessity to obtain long time-series all-weather LST from 2000 to 2021, especially during the temporal gaps of MODIS between 2000 and 2002) is well presented and introduced. The objective for work is laid out clearly in the introduction section.
The adopted approaches and methodologies are detailed, and sound and the supporting visual material is useful to better understand the exposed concepts. The experimental setup for the data acquisition and data processing is well structured and presented.
The results comprehensively demonstrates the data quality of TRIMS LST by comparison with reanalysis data, satellite TIR LST products, validation against in-situ LST, and quantification of the similarity between the TRIMS-Aqua LST and TRIMS-Terra LST time series during the temporal gaps.
Therefore, for final publication, I recommend the manuscript could be accepted after reviewed by editor.
Sincerely

Citation: https://doi.org/10.5194/essd-2023-27-RC1
- AC1: 'Reply on RC1', Wenbin Tang, 24 Apr 2023
  
  Thank you for your affirmation and praise for our work.
  
  We will continue to improve the method of estimating all-weather land surface temperature and continue to release the latest TRIMS LST publicly.
  
  Citation: https://doi.org/10.5194/essd-2023-27-AC1
RC2: 'Comment on essd-2023-27', Anonymous Referee #2, 28 Apr 2023

Please find the comment attached.

Citation: https://doi.org/10.5194/essd-2023-27-RC2
RC3: 'Comment on essd-2023-27', Anonymous Referee #3, 10 May 2023

Generating gap-free LST datasets have recently received increased interest. This study presented a new gap-free daily LST data set based on MODIS with unique temporal coverage (2000-2021). The manuscript was overall well written. However, additional information on data validation and comparison will better facilitate users to understand the data set and made choices.

The uniqueness of the presented data set needs to be further explained. Many gap-free LST products have been produced. The study claims that TRIMS has two additional years coverage. Datasets based on AVHRR or geostationary senor may also have coverage for 2000-2002 too. So additional justification is needed. This point is also related to the validation section. No other gap-free data were used for comparison in the current manuscript. Such comparison is very important to justify the new dataset.

Section 4.1 compared TRIMS with two reanalysis datasets in terms of spatial patterns. In addition, TRIMS was validated against in situ in a subsequent section. Validation of the two analysis datasets using in situ should be added to provide quantitative comparison between the fused data and one of its major source data (GLDAS).

MOD21 was reported to have superior data quality. Why did the study use MOD11 over MOD21?

Do the authors have the plan to extend the dataset to global coverage?

Citation: https://doi.org/10.5194/essd-2023-27-RC3

Peer review completion

AR: Author's response | RR: Referee report | ED: Editor decision | EF: Editorial file upload

AR by Wenbin Tang on behalf of the Authors (14 Aug 2023) Author's response Author's tracked changes Manuscript

ED: Referee Nomination & Report Request started (15 Aug 2023) by Di Tian

RR by Anonymous Referee #2 (20 Aug 2023)

Suggestions for revision or reasons for rejection

The authors responded to certain inquiries I raised, yet the feedback also prompted the emergence of further concerns. Kindly review my comments below for clarification.

Major issues:
(1) In the response letter, I highlighted a significant issue pertaining to the application of the Terra LST model to Aqua LST. The authors' explanation mentioned that “Firstly, the LST of the same pixel at different time points during a given day satisfies the standard conditions of similar pixels in the RTM method (Zhang et al., 2021). This forms the theoretical foundation for the conversion between Terra/MODIS LST and Aqua/MODIS LST (Li et al., 2018).” I'm struggling to grasp the meaning of "satisfied the standard condition of similar pixels in the RTM method." Could you provide a comprehensive explanation of this concept? Furthermore, it's important for the manuscript to explicitly address and rectify such conceptual flaws for future enhancements.

(2) At line 227 of the revised manuscript, I interpreted this sentence that RF models in module 2 appear to have been built individually for each pixel. This approach raises concerns about potential inefficiencies. How did you manage to ensure that all these models do not fall victim to overfitting? Additionally, the reliance on soil moisture data as input sparks concerns regarding the true significance of deep soil moisture contributions to models (even those from 40 cm deep). To bolster the validity of the input data selection, it would be beneficial to include a quantitative analysis in this regard.

(3) Figure 10 presents the temporal continuity analysis, but the inclusion of the annual cycle seems to obscure the clarity of the data series. I propose incorporating figures that focuses solely on anomaly variations after removing the annual cycles. More critically, it is advisable to specify the connection period (around March 2000 for Terra and July 2022 for Aqua) for meticulous analysis, particularly to demonstrate the absence of series disruptions at the conclusion of the filled temporal span.
Tech issues:
(1) Will this dataset keep updating, or it will stop at 2022?
(2) Line 18: remove ‘i.e.,’
(3) Line 25: typo ‘and/AATSR’
(4) Line 38: suggest replacing the citation ‘Jiang and Liu 2014’ to an actual land surface modeling study that used LST products.
(5) Line 105: The introduction section should clarify this work is the succession of Zhang et al 2021 and basic introduction of the original RTM paper should be given.
(6) Line 123: what is daily synthesis variable.
(7) Line 124: why GLASS albedo is only used for prediction period, if so, the difference of albedo products for the training and prediction period will affect the results as there are clear bias in different albedo products.
(8) Section 2.2: any resources and information for accessing the ground measurement networks?
(9) Line 211 and 212: duplicated statements, suggest double the manuscript for typos.
(10) It is good to summarize the available all-weather LST products; however, if the authors try to include such summary, a complete literature review should be done. Some products, e.g. high resolution LST over Europe (Rains et al. 2022), 2-km hourly product over the US (Jia et al. 2022), and hourly LST over China (https://data.tpdc.ac.cn/zh-hans/data/06414391-abd4-4d28-a844-bd036a0b8c55/)

References
Dong, Shengyue, et al. "A Data Fusion Method for Generating Hourly Seamless Land Surface Temperature from Himawari-8 AHI Data." Remote Sensing 14.20 (2022): 5170.
Jia, Aolin, Shunlin Liang, and Dongdong Wang. "Generating a 2-km, all-sky, hourly land surface temperature product from Advanced Baseline Imager data." Remote Sensing of Environment 278 (2022): 113105.
Rains, Dominik, et al. "High-resolution all-sky land surface temperature and net radiation over Europe." Earth System Science Data Discussions (2022): 1-22.
Zhang, Xiaodong, et al. "A practical reanalysis data and thermal infrared remote sensing data merging (RTM) method for reconstruction of a 1-km all-weather land surface temperature." Remote Sensing of Environment 260 (2021): 112437.

Hide

RR by Anonymous Referee #3 (05 Sep 2023)

ED: Reconsider after major revisions (05 Sep 2023) by Di Tian

AR by Wenbin Tang on behalf of the Authors (14 Nov 2023) Author's response Author's tracked changes Manuscript

ED: Referee Nomination & Report Request started (15 Nov 2023) by Di Tian

RR by Anonymous Referee #2 (16 Nov 2023)

ED: Publish as is (16 Nov 2023) by Di Tian

AR by Wenbin Tang on behalf of the Authors (21 Nov 2023) Manuscript

Short summary

This paper reported a daily 1 km all-weather land surface temperature (LST) dataset for Chinese land mass and surrounding areas – TRIMS LST. The results of a comprehensive evaluation show that TRIMS LST has the following special features: the longest time coverage in its class, high image quality, and good accuracy. TRIMS LST has already been released to the scientific community, and a series of its applications have been reported by the literature.