the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 18 Dec 2024
CCD-Rice: A long-term paddy rice distribution dataset in China at 30 m resolution
Abstract. As one of the most widely cultivated grain crops, paddy rice is a vital staple food in China and plays a crucial role in ensuring food security. Over the past decades, the planting area of paddy rice in China has shown substantial variability. Yet, there are no long-term high-resolution rice distribution maps in China, which hinders our ability to estimate greenhouse gas fluxes and crop production. This study developed a new optical satellite-based rice mapping method using a machine learning model and appropriate data preprocessing strategies to address the challenges of cloud contamination and missing data in optical remote sensing observations. This study produced CCD-Rice (China Crop Dataset-Rice), the first high-resolution rice distribution dataset in China from 1990 to 2016. Based on 397,414 validation samples, the overall accuracy of the distribution maps in each provincial administrative region averaged 89.89 %. Compared with 20,544 county-level statistical data, the coefficients of determination (R2) of single- and double-season rice in each year averaged 0.85 and 0.78, respectively. The distribution maps can be obtained at https://doi.org/10.57760/sciencedb.15865 (Shen et al., 2024a).
- Preprint
(4309 KB) - Metadata XML
-
Supplement
(206 KB) - BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on essd-2024-584', Anonymous Referee #1, 03 Jan 2025
I am very appreciated to give me the chance to review such an interesting article, the longer period, higher resolution, and covering the whole China. Such a valuable dataset will hugely improve the ability to estimate greenhouse gas fluxes, climate change impacts and crop production. The authors developed a machine learning model after taking appropriate data preprocessing strategies (so-called addressing the challenges of cloud contamination and missing data in optical remote sensing observations) to produce CCD-Rice (China Crop Dataset-Rice), from 1990 to 2016, with the overall accuracy of 89.89 % at a provincial administrative region, and 85% and 78% for single- and double-season rice at a county administrative scale. Their study is fallen closely within ESSD, and will potentially attract wider readerships. I am highly concerning on the issue of your samples for training and evaluating. However, I am several concerns listed bellows:
- Training samples: how many training samples were collected? Where are their locations?Is there any time-features (day, month, year etc.) labelled for these training samples. Are they isolated from your validation samples.
- Accuracy: remote sensing products are generally validated by two ways: sampling points by a confusion matrix, and identified areas by statistics books (both province and county, it is more popular at a smaller administrative scale). Please clearly showcase your results related with two validation methods, and clearly involve them into key-contents sections, such as ABSTRACT.
- Comparison: As conducting your comparison with other open products, it will be more scientific and objective to downscale your resolution into coarser one to align your product with other coarser product; or upscale your resolution into a higher resolution for comparing yours with other refined product.
- It will be better to emphasize your improvements comparing with the recent publications (Shen et al. 2023a, 2023b; Pan et al. 2021a, 2021b). If no any significant improvement has been made, I think it will be more reasonable to update the related datasets, rather than blindly pursuing the number of publications.
- Other: As you said in ABSTRACT “This study developed a new optical satellite-based rice mapping method using a machine learning model and appropriate data preprocessing strategies to address the challenges of cloud contamination and missing data in optical remote sensing observations. It will be better to remove “address the challenges of cloud contamination and missing data in optical remote sensing observations” because you have NOT input any substantive efforts to improve the image quality or solve the issue of cloud contamination and missing data. Please remove such statement throughout your paper to avoid readers’ discontent and protect yourselves.
- Carefully check the manuscript to eliminate the spelling mistakes，e.g. resent in Fig.3
Citation: https://doi.org/10.5194/essd-2024-584-RC1 -
RC2: 'Comment on essd-2024-584', Anonymous Referee #2, 25 Jan 2025
I am appreciated to have a chance to review this submission. Paddy rice is one of the world’s most important staple food crops, feeding half of the population. Also, rice fields are one of the main sources of greenhouse gas (GHG) emissions, contributing 12%–26% of global anthropogenic methane emissions. As the world’s largest rice producer, mapping the long-term dynamics of paddy rice is critical to food supply security and global climate change mitigation.
This study developed a new rice-mapping method with machine learning models to train the remote sensing images throughout the entire growing season of rice, and produce the whole cropping system (single- and double seasons). This product contributes to the China Crops Dataset (CCD)-Rice. Trained and validated with 397,414samples, the overall accuracy of the rice distribution maps at the provincial level was 89.89 %. Their study is within the scope of ESSD, and will be attractive to the readers of the community. However, I have major concerns as below:
For the issue of recent crop mapping publications (Shen et al. 2023a, Pan et al. 2021a), I suggest the authors to (1) clarify the CH4 emission and/or crop production for single- and double seasons rice, respectively. (2) If forcing and model available, the contribution of dynamics of single- and double seasons rice to the variations of CH4 emission and/or crop production during 1990-2016. I know this might be difficult due to the data availability and calculation cost. Then the authors could add sentences in the discussion to highlight the scientific value of this new rice mapping study.
Minor:
- Technical part: I suggest to emphasize the entire growing season of rice in southern China, rather than just during the transplanting period, resulting in more usable images for rice classification. Compared with previous studies, this study mapping both single-season and double season (much in southern China) rice.
- I suggest to rephrase the sentence 'previous studies have failed to achieve long-term, high-resolution rice mapping in China.' or delete it (Line398).
Citation: https://doi.org/10.5194/essd-2024-584-RC2
Data sets
CCD-Rice: A paddy rice distribution dataset in China from 1990 to 2016 at 30 m resolution Ruoque Shen, Qiongyan Peng, Xiangqian Li, Xiuzhi Chen, and Wenping Yuan https://doi.org/10.57760/sciencedb.15865
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 262 | 50 | 8 | 320 | 21 | 4 | 3 |
- HTML: 262
- PDF: 50
- XML: 8
- Total: 320
- Supplement: 21
- BibTeX: 4
- EndNote: 3
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1