Site-Independent Mapping of Clay Content in Vineyard Soils via Mobile Proximal Gamma-Ray Spectrometry and Machine Learning Calibrations
Abstract
:1. Introduction
2. Materials and Methods
2.1. Characterization of the Study Sites
2.2. Determination of Ground Truth Data
2.3. Stop-and-Go Gamma Measurements
2.4. Spatial Predictions of on-the-go Gamma Measurements
2.5. Tested Machine Learning Methods
2.6. Machine Learning Model Calibration
3. Results and Discussion
3.1. Variability of Gamma Features and Geopedological Setting
3.2. Calibrating Site-Independent Prediction Models for Soil Clay Content
3.3. Model Applicability for the Site-Specific Prediction of Clay Content
3.4. Applicability of Site-Independent RF Calibrations and on-the-go Measurements for Mapping the Clay Content
4. Conclusions
Author Contributions
Funding
Institutional Review Board Statement
Informed Consent Statement
Data Availability Statement
Acknowledgments
Conflicts of Interest
References
- Wiesmeier, M.; Urbanski, L.; Hobley, E.; Lang, B.; von Lützow, M.; Marin-Spiotta, E.; van Wesemael, B.; Rabot, E.; Ließ, M.; Garcia-Franco, N.; et al. Soil organic carbon storage as a key function of soils—A review of drivers and indicators at various scales. Geoderma 2019, 333, 149–162. [Google Scholar] [CrossRef]
- Schneider, N.; Islam, M.; Wehrle, R.; Pätzold, S.; Brüggemann, N.; Töpfer, R.; Herzog, K. Deep incorporation of organic amendments into soils of a ‘Calardis Musqué’ vineyard: Effects on greenhouse gas emissions, vine vigor, and grape quality. Front. Plant Sci. 2023, 14, 1253458. [Google Scholar] [CrossRef] [PubMed]
- Patzold, S.; Mertens, F.M.; Bornemann, L.; Koleczek, B.; Franke, J.; Feilhauer, H.; Welp, G. Soil Heterogeneity at the Field Scale: A Challenge for Precision Crop Protection. Precis. Agric. 2008, 9, 367–390. [Google Scholar] [CrossRef]
- Bramley, R.G.V.; Ouzman, J.; Boss, P.K. Variation in vine vigour, grape yield and vineyard soils and topography as indicators of variation in the chemical composition of grapes, wine and wine sensory attributes. Aust. J. Grape Wine Res. 2011, 17, 217–229. [Google Scholar] [CrossRef]
- Fayolle, E.; Follain, S.; Marchal, P.; Chéry, P.; Colin, F. Identification of environmental factors controlling wine quality: A case study in Saint-Emilion Grand Cru appellation, France. Sci. Total Environ. 2019, 694, 133718. [Google Scholar] [CrossRef] [PubMed]
- Viscarra Rossel, R.A.; McBratney, A.B.; Minasny, B. (Eds.) Proximal Soil Sensing; Springer: Dordrecht, The Netherlands, 2010. [Google Scholar]
- Kuang, B.; Mahmood, H.S.; Quraishi, M.Z.; Hoogmoed, W.B.; Mouazen, A.M.; van Henten, E. Sensing soil properties in the laboratory, in situ, and on-line. Adv. Agron. 2012, 114, 155–223. [Google Scholar] [CrossRef]
- Cook, S.E.; Corner, R.J.; Groves, P.R.; Grealish, G.J. Use of airborne gamma radiometric data for soil mapping. Soil Res. 1996, 34, 183. [Google Scholar] [CrossRef]
- International Atomic Energy Agency. Guidelines for Radioelement Mapping Using Gamma Ray Spectrometry Data; IAEA-TECDOC-1363; IAEA: Vienna, Austria, 2003; 173p. [Google Scholar]
- Priori, S.; Bianconi, N.; Costantini, E.A. Can γ-radiometrics predict soil textural data and stoniness in different parent materials? A comparison of two machine-learning methods. Geoderma 2014, 226–227, 354–364. [Google Scholar] [CrossRef]
- Heggemann, T.; Welp, G.; Amelung, W.; Angst, G.; Franz, S.O.; Koszinski, S.; Schmidt, K.; Pätzold, S. Proximal gamma-ray spectrometry for site-independent in situ prediction of soil texture on ten heterogeneous fields in Germany using support vector machines. Soil Till. Res. 2017, 168, 99–109. [Google Scholar] [CrossRef]
- Pätzold, S.; Leenen, M.; Heggemann, T. Proximal mobile gamma spectrometry as tool for precision farming and field experimentation. Soil Syst. 2020, 4, 31. [Google Scholar] [CrossRef]
- Van der Klooster, E.; van Egmond, F.M.; Sonneveld, M.P.W. Mapping soil clay contents in Dutch marine districts using gamma-ray spectrometry. Eur. J. Soil Sci. 2011, 62, 743–753. [Google Scholar] [CrossRef]
- Piikki, K.; Söderström, M. Digital soil mapping of arable land in Sweden—Validation of performance at multiple scales. Geoderma 2019, 352, 342–350. [Google Scholar] [CrossRef]
- Mahmood, H.F.; Hoogmoed, W.B.; van Henten, E.J. Proximal gamma-ray spectroscopy to predict soil properties using windows and full-spectrum analysis methods. Sensors 2013, 13, 16263–16280. [Google Scholar] [CrossRef] [PubMed]
- Van Egmond, F.M.; Loonstra, E.H.; Limburg, J. Gamma ray sensor for topsoil mapping: The Mole. In Proximal Soil Sensing; Viscarra Rossel, R.A., McBratney, A.B., Minasny, B., Eds.; Springer: Dordrecht, The Netherlands, 2010; pp. 323–332. [Google Scholar]
- Wadoux, A.M.-C.; Minasny, B.; McBratney, A.B. Machine learning for digital soil mapping: Applications, challenges and suggested solutions. Earth-Sci. Rev. 2020, 210, 103359. [Google Scholar] [CrossRef]
- Padarian, J.; Minasny, B.; McBratney, A.B. Using deep learning to predict soil properties from regional spectral data. Geoderma Reg. 2019, 16, e00198. [Google Scholar] [CrossRef]
- Robinson, T.P.; Metternicht, G. Testing the performance of spatial interpolation techniques for mapping soil properties. Comput. Electron. Agric. 2006, 50, 97–108. [Google Scholar] [CrossRef]
- McBratney, A.B.; Mendonça Santos, M.; Minasny, B. On digital soil mapping. Geoderma 2003, 117, 3–52. [Google Scholar] [CrossRef]
- Loonstra, E.H.; van Egmond, F.M. On-The-Go Measurement of Soil Gamma Radiation; The Soil Company: Groningen, The Netherlands, 2009; Available online: https://www.researchgate.net/profile/eddie-loonstra/publication/267379912_on-the-go_measurement_of_soil_gamma_radiation/links/563212da08ae13bc6c36ca17/on-the-go-measurement-of-soil-gamma-radiation.pdf (accessed on 8 July 2024).
- Van der Veeke, S.; Limburg, H.J.; Koomans, R.; Söderström, M.; De Waal, S.N.; Van der Graaf, E.R. Footprint and high corrections for UAV-borne gamma-ray spectrometry studies. J. Environ. Radioact. 2021, 231, 106545. [Google Scholar] [CrossRef] [PubMed]
- ISO 11277; Soil Quality: Determination of Particle Size Distribution in Mineral Soil Material, Method by Sieving and Sedimentation. International Organization for Standardization: Geneva, Switzerland, 1998.
- Hastie, T.; Tibshirani, R.; Friedman, J.H. The Elements of Statistical Learning: Data Mining, Inference, and Prediction, 2nd ed.; Springer: New York, NY, USA, 2009. [Google Scholar]
- Heung, B.; Ho, H.C.; Zhang, J.; Knudby, A.; Bulmer, C.E.; Schmidt, M. An overview and comparison of machine-learning techniques for classification purposes in digital soil mapping. Geoderma 2016, 265, 62–77. [Google Scholar] [CrossRef]
- MacKay, D.J.C. A Practical Bayesian framework for backpropagation Networks. Neural Comput. 1992, 4, 448–472. [Google Scholar] [CrossRef]
- Foresee, D.F.; Hagan, M.T. Gauss-Newton approximation to bayesian learning. In Proceedings of the international Conference on Neuronal Networks, Houston, TX, USA, 12 June 1997. [Google Scholar] [CrossRef]
- Pal, M. Random forest classifier for remote sensing classification. Int. J. Remote Sens. 2005, 26, 217–222. [Google Scholar] [CrossRef]
- Mountrakis, G.; Im, J.; Ogole, C. Support vector machines in remote sensing: A review. ISPRS J. Photogramm. Remote Sens. 2011, 66, 247–259. [Google Scholar] [CrossRef]
- Lipkowitz, K.B.; Cundari, T.R. Reviews in Computational Chemistry; John Wiley & Sons: Hoboken, NJ, USA, 2007; Volume 23. [Google Scholar]
- Kuhn, M. Building predictive models in R using the caret package. J. Stat. Softw. 2008, 28, 1–26. [Google Scholar] [CrossRef]
- Bellon-Maurel, V.; Fernandez-Ahumada, E.; Palagos, B.; Roger, J.-M.; McBratney, A. Critical review of chemometric indicators commonly used for assessing the quality of the prediction of soil attributes by NIR spectroscopy. TrAC Trends Anal. Chem. 2010, 29, 1073–1081. [Google Scholar] [CrossRef]
- Ludwig, B.; Murugan, R.; Parama, V.R.R.; Vohland, M. Accuracy of estimating soil properties with mid-infrared spectroscopy: Implications of different chemometric approaches and software packages related to calibration sample size. Soil Sci. Soc. Am. J. 2019, 83, 1542–1552. [Google Scholar] [CrossRef]
- Coulouma, G.; Caner, L.; Loonstra, E.H.; Lagacherie, P. Analysing the proximal gamma radiometry in contrasting Mediterranean landscapes: Towards a regional prediction of clay content. Geoderma 2016, 266, 127–135. [Google Scholar] [CrossRef]
- Wehrle, R.; Pätzold, S. Comparing machine learning approaches for the prediction of clay content via proximal gamma spectrometry under varying geopedological conditions. In Book of Abstracts (Posters) of the 14th European Conference on Precision Agriculture, Bologna, Italy, 2–6 July 2023. Available online: https://www.researchgate.net/publication/382183296_P28_-Comparing_machine_learning_approaches_for_the_prediction_of_clay_content_via_proximal_gamma_spectrometry_under_varying_geopedological_conditions#fullTextFileContent (accessed on 8 July 2024).
- Jordan, M.I.; Mitchell, T.M. Machine learning: Trends, perspectives, and prospects. Science 2015, 349, 255–260. [Google Scholar] [CrossRef] [PubMed]
- Zhang, M.; Shi, W.; Xu, Z. Systematic comparison of five machine-learning models in classification and interpolation of soil particle size fractions using different transformed data. Hydrol. Earth Syst. Sci. 2020, 24, 2505–2526. [Google Scholar] [CrossRef]
- Cécillon, L.; Barthès, B.G.; Gomez, C.; Ertlen, D.; Genot, V.; Hedde, M.; Stevens, A.; Brun, J.J. Assessment and monitoring of soil quality using near-infrared reflectance spectroscopy (NIRS). Eur. J. Soil Sci. 2009, 60, 770–784. [Google Scholar] [CrossRef]
Site | Wine Growing Region | Geopedological Setting |
---|---|---|
Kanz | Saar | Devonian sand, silt, and clay stones |
Leiw H | Moselle | Pleistocene fluvial deposits from Moselle middle terrace |
Leiw K | Moselle | Fluvial deposits over gravel from the Moselle middle terrace |
Rupp | Palatinate | Pleistocene fluvial deposits from sandy Rotliegend to Bunter Sandstone sedimentary rock |
Sieb B | Palatinate | Pleistocene loess |
Sieb N | Palatinate | Pleistocene periglacial slope deposits (ppsd) mainly from loamy to clayey substrates from Keuper/Bunter Sandstone sedimentary rock. Artificial Loess deposits in northern part of the field |
Spre B | Rhine-Hesse | Oligocene marl |
Spre N | Rhine-Hesse | Oligocene marl |
Site, Sample Number, Field Size | Stat. | Clay [g kg−1] | TC [cps] | K-40 [cps] | Th-232 [cps] | Th/K—Ratio |
---|---|---|---|---|---|---|
Kanz | Min. | 117 | 1187 | 173.8 | 30 | 20.8 |
n = 18 | Max. | 181 | 1690 | 258.4 | 48.7 | 31.4 |
ca. 1 ha | Mean | 148 | 1482 | 222.6 | 42.4 | 27 |
coeff. var. 1 | 12.1 | 9.81 | 10.7 | 12.7 | 3.87 | |
Leiw-H | Min. | 160 | 1041 | 142.7 | 30.7 | 20.5 |
n = 20 | Max. | 389 | 1475 | 214.2 | 42.2 | 29.3 |
ca. 1 ha | Mean | 231 | 1339 | 184.8 | 38.3 | 27.3 |
coeff. var. 1 | 24 | 7.53 | 8.44 | 7.74 | 3.51 | |
Leiw-K | Min. | 169 | 1369 | 195.4 | 39.9 | 25.7 |
n = 10 | Max. | 239 | 1648 | 240.7 | 47.7 | 33.2 |
ca. 0.5 ha | Mean | 201 | 1545 | 221.1 | 45 | 29.9 |
coeff. var. 1 | 9.52 | 5.16 | 6.84 | 5.94 | 3.85 | |
Rupp | Min. | 62.1 | 792 | 129.2 | 15.8 | 13.8 |
n = 21 | Max. | 108 | 857 | 141.2 | 17.7 | 16 |
ca. 0.4 ha | Mean | 80.2 | 822 | 135.6 | 16.5 | 15 |
coeff. var. 1 | 16.9 | 1.87 | 2.51 | 3.61 | 4.07 | |
Sieb B | Min. | 75 | 1358 | 182.5 | 36.4 | 31.1 |
n = 40 | Max. | 316 | 1488 | 203.9 | 42.3 | 34.3 |
ca. 0.3 ha | Mean | 161 | 1433 | 195 | 39.8 | 33 |
coeff. var. 1 | 31.6 | 2.58 | 3.02 | 3.67 | 3 | |
Sieb N | Min. | 101 | 1302 | 174.5 | 34.2 | 0.11 |
n = 42 | Max. | 430 | 1890 | 351 | 39.8 | 0.21 |
ca. 0.4 ha | Mean | 309 | 1590 | 259.5 | 27.9 | 0.15 |
coeff. var. 1 | 32.4 | 12.7 | 22.5 | 3.58 | 21.1 | |
Spre B | Min. | 113 | 895 | 114.5 | 22 | 20.1 |
n = 53 | Max. | 647 | 1027 | 145.7 | 29.2 | 26.9 |
ca. 1 ha | Mean | 505 | 970 | 130.1 | 25.4 | 23.8 |
coeff. var. 1 | 22.5 | 3.59 | 6.68 | 7.19 | 6.46 | |
Spre N | Min. | 277 | 1026 | 139.4 | 27.4 | 21.9 |
n = 31 | Max. | 629 | 1134 | 163.5 | 31.6 | 28.4 |
ca. 0.6 ha | Mean | 483 | 1093 | 152.7 | 29.8 | 25 |
coeff. var. 1 | 18.5 | 2.21 | 4.37 | 4.05 | 5.07 |
Model | Cross-Validation | Test Set Validation | ||||
---|---|---|---|---|---|---|
RMSECV [g kg−1] | R2CV | RPIQCV | RMSEPr [g kg−1] | R2Pr | RPIQPr | |
SVM | 62.3 | 0.87 | 4.77 | 80.8 | 0.80 | 4.02 |
KNN | 78.5 | 0.79 | 3.78 | 91.7 | 0.75 | 3.54 |
BNN | 72.2 | 0.83 | 4.11 | 93.2 | 0.74 | 3.48 |
RF | 36.8 | 0.96 | 8.65 | 57.6 | 0.87 | 4.64 |
Site | RMSE [g kg−1] | R2 | RPIQ |
---|---|---|---|
Kanz | 27.9 | 0.4 | 0.61 |
Leiw H | 38.6 | 0.55 | 1.69 |
Leiw K | 15.9 | 0.42 | 1.12 |
Rupp | 5.5 | 0.92 | 4.39 |
Sieb B | 28.9 | 0.89 | 2.06 |
Sieb N | 19.0 | 0.97 | 9.24 |
Spre B | 51.1 | 0.88 | 2.27 |
Spre N | 52.4 | 0.79 | 1.72 |
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content. |
© 2024 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).
Share and Cite
Wehrle, R.; Pätzold, S. Site-Independent Mapping of Clay Content in Vineyard Soils via Mobile Proximal Gamma-Ray Spectrometry and Machine Learning Calibrations. Sensors 2024, 24, 4528. https://doi.org/10.3390/s24144528
Wehrle R, Pätzold S. Site-Independent Mapping of Clay Content in Vineyard Soils via Mobile Proximal Gamma-Ray Spectrometry and Machine Learning Calibrations. Sensors. 2024; 24(14):4528. https://doi.org/10.3390/s24144528
Chicago/Turabian StyleWehrle, Ralf, and Stefan Pätzold. 2024. "Site-Independent Mapping of Clay Content in Vineyard Soils via Mobile Proximal Gamma-Ray Spectrometry and Machine Learning Calibrations" Sensors 24, no. 14: 4528. https://doi.org/10.3390/s24144528
APA StyleWehrle, R., & Pätzold, S. (2024). Site-Independent Mapping of Clay Content in Vineyard Soils via Mobile Proximal Gamma-Ray Spectrometry and Machine Learning Calibrations. Sensors, 24(14), 4528. https://doi.org/10.3390/s24144528