the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Addressing soil data needs and data-gaps in catchment scale environmental modelling: the European perspective
Abstract. To effectively guide agricultural management planning strategies and policy, it is important to simulate water quantity and quality patterns and quantify the impact of land use and climate change on underlying processes. Environmental models that depict alterations in surface and groundwater quality and quantity at a catchment scale require substantial input, particularly concerning movement and retention in the unsaturated zone. Over the past few decades, numerous soil information sources, containing structured data on diverse basic and advanced soil parameters, alongside innovative solutions to estimate missing soil data, have become increasingly available. This study aims to: i) catalogue open-source soil datasets and pedotransfer functions (PTFs) applicable in simulation studies across European catchments, ii) evaluate the performance of selected PTFs and iii) present compiled R scripts proposing estimation solutions to address soil physical, hydraulic, and chemical soil data needs and gaps in catchment-scale environmental modelling in Europe. Our focus encompassed basic soil properties, bulk density, porosity, albedo, soil erodibility factor, field capacity, wilting point, available water capacity, saturated hydraulic conductivity, and phosphorus content. We aim to recommend widely supported data sources and pioneering prediction methods that maintain physical consistency, and present them through streamlined workflows.
- Preprint
(2619 KB) - Metadata XML
-
Supplement
(1507 KB) - BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on egusphere-2023-3104', Anonymous Referee #1, 25 Mar 2024
Thanks to the authors for putting in review this topic.
My major concerns:
L40. Why information on hydraulic properties is often lacking. Particularly, now that the EU Commission proposes the Soil Monitoring Law and such attributes can be important for soil health. By the way, you did not mention anything in your manuscript about soil health. How those attributes are linked to Soil health and how to be used to estimate soil health?
Table 1 is not easily readable as proposed. I would propose to have a landscape format and add also a column as Reference. Practically, transfer the reference from first column to the new one. Some corrections are also necessary:
- It is European Union (EU). Of course if you state Member states is similar but more precise to refer to the EU.
- In Soil hydraulic or Physical data , you can add (as there are new datasets that have not been taken into account in your analysis ; such as the Bulk density in the EU and the Global K-factor:
- https://esdac.jrc.ec.europa.eu/content/topsoil-physical-properties-europe-based-lucas-topsoil-data *** Clay, silt and sand content; coarse fragments; bulk density; USDA soil textural class; available water capacity. Resolution 500m. as in Ballabio et al., 2016https://esdac.jrc.ec.europa.eu/content/chemical-properties-european-scale-based-lucas-topsoil-data *** pH, pH (CaCl), Cation Exchange Capacity (CEC), Calcium carbonates (CaCO3), C:N ratio, Nitrogen (N), Phosphorus (P) and Potassium (K) as in Ballabio et al., 2019
https://esdac.jrc.ec.europa.eu/content/soil-bulk-density-europe ****** Soil Bulk density in Europe as in Panagos et al., 2024
https://esdac.jrc.ec.europa.eu/content/global-soil-erodibility ***** The Global K-factor of Gupta et al. 2024
Those are different that LUCAS. You mentioned LUCAS but those datasets are derived through LUCAS with Machine learning.
L116-117: in this place and other places of the manuscript. You mentioned this sentence but you have to admit that also application of PTF has a huge uncertainty and it not proper for all different pedo-climatic regions of Europe. EU is so diverse that the just one PTR is not the valid approach for the whole EU. Your statements such as this one as so strong and negative towards other estimations or assessments. In general, I would suggest a more multi-model approach where assessments based on machine learning or interpolations can be compared or assessed together with assessments of PTF.
The evaluation of your results takes place using the EU-HYDI . This database as you state is not publicly available. This is really odd and not transparent. Others do European assessments (Soil Grids, ESDAC) or Global assessments but the point data (LUCAS, global point data) are available. Therefore, the approaches are transparent and everybody can test them.
In Bulk density, the BD of Hollis is not so simple. Hollis has proposed different PTF per different land uses. Therefore, please pay attention in this use. In addition, as mentioned before, you ignored the recent public assessment of EU Bulk density with 6,000 points available (to download).
For K-factor, the most used function is the by Wischmeier and Smith (1978) and Renard et al. (1997) (as described in Panagos et al., in Eq. (1)).
You use a different equation and then you try to compare your results with the ones which have used the Renard equation. This is a little bit odd.In 2.4, for P it is not only the fertilization which plays a role. The available P in soils is a combination of P inputs (Fertilizers, manure, atmospheric deposition, chemical weathering) and outputs (plant uptake, plant residues , erosion). Therefore P level is not influenced only by fertilization. Please be careful and change as appropriate!
In soil chemical parameters, authors do not explain why not N and K?
In section 3.1, the performance of BD PTF is not valid as I explained my problematic for the Hollis eq (which is not used properly).
In addition, why you do not test the PTFs against the LUCAS 6000 measured data which are publicly available?
The problematics on 3.3 have also described above as your results tend to compare non comparable stuff (different equations used!!!).
In 618-619: you refer to something that it is too obvious. IF there are local data, of course they are better. The case is how to cover the data gaps in case local data are not available? That is why I have proposed a multi-model or multi-data source assessment?
Similar in your conclusions L653. It is too obvious!
L675-685: you cannot propose this as the only way forward without making available your reference dataset (EU-HYDI)!!!
Other issues:
L16: which are the underlying processes?
L28-29: why there an significant increase of available datasets?
L159: Not only different methods but also through different ISO protocols, depths, etc and in different laboratories which sometimes is impossible to compare.!!
L325: It is nitrogenCitation: https://doi.org/10.5194/egusphere-2023-3104-RC1 - RC2: 'Comment on egusphere-2023-3104', Diana Vieira, 26 Mar 2024
Model code and software
SWATprepR: SWAT+ Input Data Preparation Package Svajunas Plunge https://doi.org/10.5281/zenodo.10167076
Map topsoil phosphorus content Brigitta Szabó and Piroska Kassai https://doi.org/10.5281/zenodo.6656537
Viewed
HTML | XML | Total | Supplement | BibTeX | EndNote | |
---|---|---|---|---|---|---|
293 | 89 | 12 | 394 | 29 | 8 | 6 |
- HTML: 293
- PDF: 89
- XML: 12
- Total: 394
- Supplement: 29
- BibTeX: 8
- EndNote: 6
Viewed (geographical distribution)
Country | # | Views | % |
---|
Total: | 0 |
HTML: | 0 |
PDF: | 0 |
XML: | 0 |
- 1