I am Wei Shangguan from School of Atmospheric Sciences, Sun Yat-sen University. Throughout my research career starting from 2005, making the best maps (at least, we tried to be the best) in Earth sciences for China and the world is a major topic. Usually, these efforts need cooperation with others. Until now, my coworkers and I have released 11 data sets at http://globalchange.bnu.edu.cn/research/data, which are freely available with a registration. We have more than 10,000 registered users from all over the world, which makes us quite proud and have a feeling of value and success. I can say now I am a successful mapper :-)
The map of the depth to bedrock (DTB) of China (https://doi.org/10.1038/s41597-019-0345-6) is one of the above efforts, which is the first one at the national scale. Before this map, my coworkers and I have already released a depth to bedrock of the world in 2017, which is still the best for the whole world. But I am not satisfied because this map still contains so much uncertainty. To make better knowledge of DTB, I guided my master student, Fapeng Yan, to map it for China. He was quite endeavored to this job. Data collection was quite boring and laborious because we needed to interpret DTB from old geological logs in the form of pictures. Most of these DTBs were interpreted by Fapeng, and I told him to do so not at once but accompanied with coding and paper reading to reduce the bore. At last, we got more than 6,000 valid observations. After data collection, we used artificial intelligence (here refers to machine learning) to map DTB from point observations to spatial distributions. It is not an automatic process though it is called artificial intelligence. The spatial modelling requires human intelligence quite a bit. I won’t tell all the difficulties but we have a failure in the application of deep learning to map DTB which is not reported in the paper. The publication process is also long. This work has been rejected by another journal with two rounds of major revision and another instant rejection. As a result, this paper is honored to be accepted by Scientific Data now (luckily) when it is about one and half year later since its first submission to another journal. I can say this map was the best DTB map of China when its first version was produced one and half year ago (evidences provided in the paper). And It will be the best DTB map of China for some time. Of course, I am very thankful for some helpful comments from the reviewers which helped out a better map and paper, but not hateful for some tough comments that has no value in making a better map. I know this procedure is normal but it wastes our time and thus the time to use it, especially for guys like me not quite good at writing an English scientific paper. Anyway, I insist that for data publication, it is not necessary for reviewers and editors to care too much about things that have no value or marginal value to improve it and what we need is a good and useful data set.
Why do we need to know DTB? The direct purpose is for the use in the land surface models, because I am a member of land surface modeling team at Sun Yat-sen University (http://globalchange.bnu.edu.cn). Land surface model is one of the important component of Earth system models and climate/weather models. The lower boundary of land surface model is DTB, which has influences on water, energy and bio-geochemistry (e.g. carbon) cycles. DTB is also the lower boundary of the so-called Critical Zone of the Earth and you can imagine a good map of DTB will definitely benefit related Earth science research and engineering such as hydrology, ecology and agriculture. Until recent years, this lower boundary remains unknown as a map. The potential usages are good and wide but need intelligent work from different experts knowing how to use it. Our best reward is that people can make best use of this map. Maps can always be better and we will continue to make maps and improve them, just like our ancestors who explored the unknown world and making maps.