the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Global agricultural lands in the year 2015
Abstract. While there are many global geospatial datasets representing the extent of agriculture, they predominantly represent croplands. Only a couple of global data products represent the full global agricultural footprint, including pastures. Our own research team’s most recent complete publicly available agricultural land cover dataset, including both croplands and pastures, represent circa 2000. These data, distributed on a graticule of 5 arcminutes (~10 km2 at the equator), have been integrated into a considerable number and diversity of research studies, modeling, data science and media applications. Further, users of these data have been interested in them for studying a variety of issues such as land use, food security, climate change and biodiversity loss. Here we present an updated dataset on the global distribution of agricultural lands (cropland and pasture) circa 2015 (15 years on since the initial study). Past studies that have constructed such datasets have been one-off exercises that have been infrequently repeated due to the amount of effort required. Therefore, in this work, we developed a transparent and reproducible approach to update our data product while also enabling easier reproduction of future datasets. We distribute our 2015 product at the same resolution and formats as the prior product, and accompany it with a full set of replicable code and data for reconstruction. In this article we explain how the data was constructed, with links to the permanent DOIs where the data can be readily downloaded by the user community (Mehrabi et al. 2024; DOI: 10.5281/zenodo.11540554).
- Preprint
(3103 KB) - Metadata XML
- BibTeX
- EndNote
Status: open (until 20 Sep 2024)
-
RC1: 'Comment on essd-2024-279', Lindsey Sloat, 28 Aug 2024
reply
General comments:
This manuscript presents an important update to a widely used global dataset of agricultural lands (including both cropland and pasture) for circa 2015. As a researcher working in this space, I can confidently say that these data are needed and will be widely used. As an anecdote, I have interacted with many researchers in the academic, NGO, and private space that still use the original data (circa 2000) in their analyses, as no other datasets exist that are global, comprehensive in both crop and pasture percent area, and are served at a resolution readily useable for global models and other land use and ghg accounting metrics.
This manuscript is being submitted in 2024 for maps that reflect 2015 and by the time of publication may be 10 years out of date. While it would be great to have something more updated, this delay is likely a reflection of the time it takes to receive and process census and survey data, create and document reproducible code, make comparisons with other global maps, and the general nature of academic data production. While other higher-resolution and more updated datasets have come out and will continue to come out, these 2015 data will remain very useful for the reasons stated above. Furthermore, future updates could be faster as the authors seem to have taken great care to create a reproducible pipeline for updating future versions of the maps, which is a great service to the community and will ensure reproducibility, trust, uptake, and longevity.
The methods used in the map production are tested and sound, as far as I’m aware. The bias-correction steps and post-processing methods, including pycnophylactic interpolation, seem appropriate. Because the data production pipeline is open, others can assess the impact that these steps have on the final product. The authors have made the proper statements about the appropriate use of the data (“…these data are intended for use in global modeling studies… This update is for users that require global data that covers comprehensive cropland and pasture definitions and is numerically consistent between land use estimates”). There are possibly some idiosyncrasies and missing areas in the maps, which I mention below. I make one suggestion to either adjust the latitude/GDD mask for pastures to better reflect reality or at least to document the extent of those missing areas when comparing to HYDE and HILDA+.
I am not an expert in the validation of geospatial data products so I would defer to other reviewers on this topic. I understand that the modern best practice involves an independent visual inspection, but it also seems to me that it would be extremely challenging (or not possible?) to do that on a percent area product at this resolution.
Overall, the manuscript and accompanying maps and code represent a valuable contribution to the field of global agricultural mapping.
Specific Comments
- I’m glad that the data production pipeline supports the production of maps that are not aligned to FAOStat, given some of the known inaccuracies. Will those maps receive a separate peer-reviewed publication? If not, I would encourage the authors to consider presenting them here in the supplemental material. I believe these maps would be valuable but far less used if not peer-reviewed. It doesn’t seem like it would have to add much length to the manuscript text if the production is just a branch of the current pipeline. I should clarify that this comment is more of a personal recommendation that I think would strengthen the paper and make the non-FAO matched product more useable. I don’t think it should be taken as a pre-condition for publication as the current version does stand on its own.
- It would be useful to define what is a pasture or cropland area for this map as well as some heuristics about what we should expect to find in these areas. Are you adopting the FAO definitions of arable land and permanent meadows and pastures and calling it cropland and pasture? Is there a defined allowable fallow period to be considered cropland? Having an explicit definition will make it easier for others to understand if this map is suitable for their use case and also help to “map” the differences between this product and others. If the definition is in the old paper I think it’s worth repeating.
- The Northern latitude or GDD mask may be too strict (I think more for pastures than croplands) as it seems to be masking out some areas that should be considered agriculture such as the UK and Northern Ireland, Fennoscandia, and Iceland. You could compare to results here (https://link.springer.com/article/10.1007/s10980-024-01810-6) or here (https://link.springer.com/article/10.1038/s41598-022-20095-w?fromPaywallRec=false). Additionally, in Figure 5 it looks like the latitude/GDD mask was applied to HYDE and HILDA as well before making the comparison but it would be useful to see how much pasture area was included in those datasets that’s excluded from this one. I would recommend either reconsidering these constraints or at least quantifying the impact by showing how much area is left out compared to other products when this constraint is applied.
- Starting on line 240 it’s confusing to understand how you’re treating Australia and why. For example, you mention masking grazing in the cropland maps but how are the Australian grazing areas defined? Maybe it’s the Abares reference but you only mention pasture and not grazing as a term when you describe it. It would be nice to give some additional details and justifications for why an aridity mask is only used in Australia. Perhaps a summary of the points made in the original paper would help this section flow better and stand alone.
- Line 270. You could consider adding some additional data on top of Geo-wiki or Potopov to cover some of the missing perennial crops. See here for potential data sources (https://www.wri.org/research/spatial-database-planted-trees-sdpt-version-2).
Technical corrections
- Line 210 – I was unfamiliar with the term “stride” but if it’s commonly understood there is no need to change it.
- Figure 3. I think the axes should have units of percent
- 271 – Geo-Wiki is hyphenated
- 290 – I think there must be a typo here and it’s actually 2012-2015 and 2016-2019.
- 340 – Table 2 caption should read “… area estimates”
All the best to the authors for their hard work on this,
Lindsey Sloat
Citation: https://doi.org/10.5194/essd-2024-279-RC1
Data sets
Geospatial database of global agricultural lands in the year 2015 Zia Mehrabi, Kaitai Tong, Julie Fortin, Radost Stanimirova, Mark Friedl, and Navin Ramankutty https://doi.org/10.5281/zenodo.11540554
Viewed
HTML | XML | Total | BibTeX | EndNote | |
---|---|---|---|---|---|
249 | 119 | 14 | 382 | 8 | 7 |
- HTML: 249
- PDF: 119
- XML: 14
- Total: 382
- BibTeX: 8
- EndNote: 7
Viewed (geographical distribution)
Country | # | Views | % |
---|
Total: | 0 |
HTML: | 0 |
PDF: | 0 |
XML: | 0 |
- 1