Articles | Volume 16, issue 8
https://doi.org/10.5194/gmd-16-2149-2023
© Author(s) 2023. This work is distributed under
the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
https://doi.org/10.5194/gmd-16-2149-2023
© Author(s) 2023. This work is distributed under
the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Causal deep learning models for studying the Earth system
Institute of Bio- and Geosciences, Agrosphere (IBG-3), Forschungszentrum Jülich, 52425 Jülich, Germany
Center for High-Performance Scientific Computing in Terrestrial Systems, Geoverbund ABC/J, 52425 Jülich, Germany
Stefan Kollet
Institute of Bio- and Geosciences, Agrosphere (IBG-3), Forschungszentrum Jülich, 52425 Jülich, Germany
Center for High-Performance Scientific Computing in Terrestrial Systems, Geoverbund ABC/J, 52425 Jülich, Germany
Jochen Garcke
Fraunhofer SCAI, 53757 Sankt Augustin, Germany
Institut für Numerische Simulation, Universität Bonn, 53115 Bonn, Germany
Related authors
No articles found.
Chen Yang, Aoqi Sun, Shupeng Zhang, Yongjiu Dai, Stefan Kollet, and Reed Maxwell
EGUsphere, https://doi.org/10.5194/egusphere-2025-3935, https://doi.org/10.5194/egusphere-2025-3935, 2025
This preprint is open for discussion and under review for Geoscientific Model Development (GMD).
Short summary
Short summary
Groundwater strongly influences how water and energy move between land and air, yet most large-scale climate and Earth system models treat it too simply. We reviewed 20 years of work combining a detailed groundwater model, ParFlow, with land surface models, showing ways groundwater shapes energy and water cycles. We also updated this model link, improving its performance, and proposed a flexible framework to support future advances.
Robert Reinecke, Annemarie Bäthge, Ricarda Dietrich, Sebastian Gnann, Simon N. Gosling, Danielle Grogan, Andreas Hartmann, Stefan Kollet, Rohini Kumar, Richard Lammers, Sida Liu, Yan Liu, Nils Moosdorf, Bibi Naz, Sara Nazari, Chibuike Orazulike, Yadu Pokhrel, Jacob Schewe, Mikhail Smilovic, Maryna Strokal, Yoshihide Wada, Shan Zuidema, and Inge de Graaf
EGUsphere, https://doi.org/10.5194/egusphere-2025-1181, https://doi.org/10.5194/egusphere-2025-1181, 2025
Short summary
Short summary
Here we describe a collaborative effort to improve predictions of how climate change will affect groundwater. The ISIMIP groundwater sector combines multiple global groundwater models to capture a range of possible outcomes and reduce uncertainty. Initial comparisons reveal significant differences between models in key metrics like water table depth and recharge rates, highlighting the need for structured model intercomparisons.
Bamidele Oloruntoba, Stefan Kollet, Carsten Montzka, Harry Vereecken, and Harrie-Jan Hendricks Franssen
Hydrol. Earth Syst. Sci., 29, 1659–1683, https://doi.org/10.5194/hess-29-1659-2025, https://doi.org/10.5194/hess-29-1659-2025, 2025
Short summary
Short summary
We studied how soil and weather data affect land model simulations over Africa. By combining soil data processed in different ways with weather data of varying time intervals, we found that weather inputs had a greater impact on water processes than soil data type. However, the way soil data were processed became crucial when paired with high-frequency weather inputs, showing that detailed weather data can improve local and regional predictions of how water moves and interacts with the land.
Elena Xoplaki, Florian Ellsäßer, Jens Grieger, Katrin M. Nissen, Joaquim G. Pinto, Markus Augenstein, Ting-Chen Chen, Hendrik Feldmann, Petra Friederichs, Daniel Gliksman, Laura Goulier, Karsten Haustein, Jens Heinke, Lisa Jach, Florian Knutzen, Stefan Kollet, Jürg Luterbacher, Niklas Luther, Susanna Mohr, Christoph Mudersbach, Christoph Müller, Efi Rousi, Felix Simon, Laura Suarez-Gutierrez, Svenja Szemkus, Sara M. Vallejo-Bernal, Odysseas Vlachopoulos, and Frederik Wolf
Nat. Hazards Earth Syst. Sci., 25, 541–564, https://doi.org/10.5194/nhess-25-541-2025, https://doi.org/10.5194/nhess-25-541-2025, 2025
Short summary
Short summary
Europe frequently experiences compound events, with major impacts. We investigate these events’ interactions, characteristics, and changes over time, focusing on socio-economic impacts in Germany and central Europe. Highlighting 2018’s extreme events, this study reveals impacts on water, agriculture, and forests and stresses the need for impact-focused definitions and better future risk quantification to support adaptation planning.
Florian Knutzen, Paul Averbeck, Caterina Barrasso, Laurens M. Bouwer, Barry Gardiner, José M. Grünzweig, Sabine Hänel, Karsten Haustein, Marius Rohde Johannessen, Stefan Kollet, Mortimer M. Müller, Joni-Pekka Pietikäinen, Karolina Pietras-Couffignal, Joaquim G. Pinto, Diana Rechid, Efi Rousi, Ana Russo, Laura Suarez-Gutierrez, Sarah Veit, Julian Wendler, Elena Xoplaki, and Daniel Gliksman
Nat. Hazards Earth Syst. Sci., 25, 77–117, https://doi.org/10.5194/nhess-25-77-2025, https://doi.org/10.5194/nhess-25-77-2025, 2025
Short summary
Short summary
Our research, involving 22 European scientists, investigated drought and heat impacts on forests in 2018–2022. Findings reveal that climate extremes are intensifying, with central Europe being most severely impacted. The southern region showed resilience due to historical drought exposure, while northern and Alpine areas experienced emerging or minimal impacts. The study highlights the need for region-specific strategies, improved data collection, and sustainable practices to safeguard forests.
Bjorn Stevens, Stefan Adami, Tariq Ali, Hartwig Anzt, Zafer Aslan, Sabine Attinger, Jaana Bäck, Johanna Baehr, Peter Bauer, Natacha Bernier, Bob Bishop, Hendryk Bockelmann, Sandrine Bony, Guy Brasseur, David N. Bresch, Sean Breyer, Gilbert Brunet, Pier Luigi Buttigieg, Junji Cao, Christelle Castet, Yafang Cheng, Ayantika Dey Choudhury, Deborah Coen, Susanne Crewell, Atish Dabholkar, Qing Dai, Francisco Doblas-Reyes, Dale Durran, Ayoub El Gaidi, Charlie Ewen, Eleftheria Exarchou, Veronika Eyring, Florencia Falkinhoff, David Farrell, Piers M. Forster, Ariane Frassoni, Claudia Frauen, Oliver Fuhrer, Shahzad Gani, Edwin Gerber, Debra Goldfarb, Jens Grieger, Nicolas Gruber, Wilco Hazeleger, Rolf Herken, Chris Hewitt, Torsten Hoefler, Huang-Hsiung Hsu, Daniela Jacob, Alexandra Jahn, Christian Jakob, Thomas Jung, Christopher Kadow, In-Sik Kang, Sarah Kang, Karthik Kashinath, Katharina Kleinen-von Königslöw, Daniel Klocke, Uta Kloenne, Milan Klöwer, Chihiro Kodama, Stefan Kollet, Tobias Kölling, Jenni Kontkanen, Steve Kopp, Michal Koran, Markku Kulmala, Hanna Lappalainen, Fakhria Latifi, Bryan Lawrence, June Yi Lee, Quentin Lejeun, Christian Lessig, Chao Li, Thomas Lippert, Jürg Luterbacher, Pekka Manninen, Jochem Marotzke, Satoshi Matsouoka, Charlotte Merchant, Peter Messmer, Gero Michel, Kristel Michielsen, Tomoki Miyakawa, Jens Müller, Ramsha Munir, Sandeep Narayanasetti, Ousmane Ndiaye, Carlos Nobre, Achim Oberg, Riko Oki, Tuba Özkan-Haller, Tim Palmer, Stan Posey, Andreas Prein, Odessa Primus, Mike Pritchard, Julie Pullen, Dian Putrasahan, Johannes Quaas, Krishnan Raghavan, Venkatachalam Ramaswamy, Markus Rapp, Florian Rauser, Markus Reichstein, Aromar Revi, Sonakshi Saluja, Masaki Satoh, Vera Schemann, Sebastian Schemm, Christina Schnadt Poberaj, Thomas Schulthess, Cath Senior, Jagadish Shukla, Manmeet Singh, Julia Slingo, Adam Sobel, Silvina Solman, Jenna Spitzer, Philip Stier, Thomas Stocker, Sarah Strock, Hang Su, Petteri Taalas, John Taylor, Susann Tegtmeier, Georg Teutsch, Adrian Tompkins, Uwe Ulbrich, Pier-Luigi Vidale, Chien-Ming Wu, Hao Xu, Najibullah Zaki, Laure Zanna, Tianjun Zhou, and Florian Ziemen
Earth Syst. Sci. Data, 16, 2113–2122, https://doi.org/10.5194/essd-16-2113-2024, https://doi.org/10.5194/essd-16-2113-2024, 2024
Short summary
Short summary
To manage Earth in the Anthropocene, new tools, new institutions, and new forms of international cooperation will be required. Earth Virtualization Engines is proposed as an international federation of centers of excellence to empower all people to respond to the immense and urgent challenges posed by climate change.
Liubov Poshyvailo-Strube, Niklas Wagner, Klaus Goergen, Carina Furusho-Percot, Carl Hartick, and Stefan Kollet
Earth Syst. Dynam., 15, 167–189, https://doi.org/10.5194/esd-15-167-2024, https://doi.org/10.5194/esd-15-167-2024, 2024
Short summary
Short summary
Groundwater (GW) representation is simplified in most regional climate models. Here, we introduce a unique Terrestrial Systems Modeling Platform (TSMP) dataset with an explicit representation of GW, in the context of dynamical downscaling of GCMs for climate change studies. We compare the heat events statistics of TSMP and the CORDEX ensemble. Our results show that TSMP systematically simulates fewer heat waves, and they are shorter and less intense.
Zbigniew P. Piotrowski, Jaro Hokkanen, Daniel Caviedes-Voullieme, Olaf Stein, and Stefan Kollet
EGUsphere, https://doi.org/10.5194/egusphere-2023-1079, https://doi.org/10.5194/egusphere-2023-1079, 2023
Preprint withdrawn
Short summary
Short summary
The computer programs capable of simulation of Earth system components evolve, adapting new fundamental science concepts and more observational data on more and more powerful computer hardware. Adaptation of a large scientific program to a new type of hardware is costly. In this work we propose cheap and simple but effective strategy that enable computation using graphic processing units, based on automated program code modification. This results in better resolution and/or longer predictions.
Bibi S. Naz, Wendy Sharples, Yueling Ma, Klaus Goergen, and Stefan Kollet
Geosci. Model Dev., 16, 1617–1639, https://doi.org/10.5194/gmd-16-1617-2023, https://doi.org/10.5194/gmd-16-1617-2023, 2023
Short summary
Short summary
It is challenging to apply a high-resolution integrated land surface and groundwater model over large spatial scales. In this paper, we demonstrate the application of such a model over a pan-European domain at 3 km resolution and perform an extensive evaluation of simulated water states and fluxes by comparing with in situ and satellite data. This study can serve as a benchmark and baseline for future studies of climate change impact projections and for hydrological forecasting.
Mohamed Saadi, Carina Furusho-Percot, Alexandre Belleflamme, Ju-Yu Chen, Silke Trömel, and Stefan Kollet
Nat. Hazards Earth Syst. Sci., 23, 159–177, https://doi.org/10.5194/nhess-23-159-2023, https://doi.org/10.5194/nhess-23-159-2023, 2023
Short summary
Short summary
On 14 July 2021, heavy rainfall fell over central Europe, causing considerable damage and human fatalities. We analyzed how accurate our estimates of rainfall and peak flow were for these flooding events in western Germany. We found that the rainfall estimates from radar measurements were improved by including polarimetric variables and their vertical gradients. Peak flow estimates were highly uncertain due to uncertainties in hydrological model parameters and rainfall measurements.
Bernd Schalge, Gabriele Baroni, Barbara Haese, Daniel Erdal, Gernot Geppert, Pablo Saavedra, Vincent Haefliger, Harry Vereecken, Sabine Attinger, Harald Kunstmann, Olaf A. Cirpka, Felix Ament, Stefan Kollet, Insa Neuweiler, Harrie-Jan Hendricks Franssen, and Clemens Simmer
Earth Syst. Sci. Data, 13, 4437–4464, https://doi.org/10.5194/essd-13-4437-2021, https://doi.org/10.5194/essd-13-4437-2021, 2021
Short summary
Short summary
In this study, a 9-year simulation of complete model output of a coupled atmosphere–land-surface–subsurface model on the catchment scale is discussed. We used the Neckar catchment in SW Germany as the basis of this simulation. Since the dataset includes the full model output, it is not only possible to investigate model behavior and interactions between the component models but also use it as a virtual truth for comparison of, for example, data assimilation experiments.
Yueling Ma, Carsten Montzka, Bagher Bayat, and Stefan Kollet
Hydrol. Earth Syst. Sci., 25, 3555–3575, https://doi.org/10.5194/hess-25-3555-2021, https://doi.org/10.5194/hess-25-3555-2021, 2021
Short summary
Short summary
This study utilized spatiotemporally continuous precipitation anomaly (pra) and water table depth anomaly (wtda) data from integrated hydrologic simulation results over Europe in combination with Long Short-Term Memory (LSTM) networks to capture the time-varying and time-lagged relationship between pra and wtda in order to obtain reliable models to estimate wtda at the individual pixel level.
Cited articles
Adler, B., Kalthoff, N., and Gantner, L.: Initiation of deep convection caused
by land-surface inhomogeneities in West Africa: a modelled case study,
Meteorol. Atmos. Phys., 112, 15–27,
https://doi.org/10.1007/s00703-011-0131-2, 2011. a
Barnes, E. A., Samarasinghe, S. M., Ebert-Uphoff, I., and Furtado, J. C.:
Tropospheric and Stratospheric Causal Pathways Between the MJO and NAO,
J. Geophys. Res.-Atmos., 124, 9356–9371,
https://doi.org/10.1029/2019jd031024, 2019. a
Baur, F., Keil, C., and Craig, G. C.: Soil moisture–precipitation
coupling over Central Europe: Interactions between surface anomalies at
different scales and the dynamical implication, Q. J.
Roy. Meteor. Soc., 144, 2863–2875, https://doi.org/10.1002/qj.3415, 2018. a
Dumoulin, V. and Visin, F.: A guide to convolution arithmetic for deep
learning, https://arxiv.org/abs/1603.07285 (last access: 16 April 2023), 2016. a
Ebert-Uphoff, I. and Deng, Y.: Causal discovery in the
geosciences – Using synthetic data to learn how to interpret
results, Comput. Geosci., 99, 50–60,
https://doi.org/10.1016/j.cageo.2016.10.008, 2017. a
Ebert-Uphoff, I. and Hilburn, K.: Evaluation, Tuning, and Interpretation of
Neural Networks for Working with Images in Meteorological Applications,
B. Am. Meteorol. Soc., 101, E2149–E2170,
https://doi.org/10.1175/bams-d-20-0097.1, 2020. a
Eltahir, E. A. B.: A Soil Moisture–Rainfall Feedback Mechanism: 1. Theory and
observations, Water Resour. Res., 34, 765–776,
https://doi.org/10.1029/97WR03499, 1998. a, b
Findell, K. L. and Eltahir, E. A. B.: Atmospheric Controls on Soil
Moisture–Boundary Layer Interactions. Part I: Framework
Development, J. Hydrometeorol., 4, 552–569,
https://doi.org/10.1175/1525-7541(2003)004<0552:acosml>2.0.co;2, 2003a. a
Findell, K. L. and Eltahir, E. A. B.: Atmospheric Controls on Soil
Moisture–Boundary Layer Interactions. Part II: Feedbacks within
the Continental United States, J. Hydrometeor., 4, 570–583,
https://doi.org/10.1175/1525-7541(2003)004<0570:acosml>2.0.co;2, 2003b. a
Froidevaux, P., Schlemmer, L., Schmidli, J., Langhans, W., and Schär, C.:
Influence of the Background Wind on the Local Soil
Moisture–Precipitation Feedback, J. Atmos.
Sci., 71, 782–799, https://doi.org/10.1175/jas-d-13-0180.1, 2014. a
Gagne II, D. J., Haupt, S. E., Nychka, D. W., and Thompson, G.: Interpretable
Deep Learning for Spatial Analysis of Severe Hailstorms, Mon. Weather
Rev., 147, 2827–2845, https://doi.org/10.1175/mwr-d-18-0316.1, 2019. a
Gentine, P., Holtslag, A. A. M., D'Andrea, F., and Ek, M.:
Surface and Atmospheric Controls on the Onset of Moist Convection over Land,
J. Hydrometeorol., 14, 1443–1462, https://doi.org/10.1175/jhm-d-12-0137.1,
2013. a
Gilpin, L., Bau, D., Yuan, B., Bajwa, A., Specter, M., and Kagal, L.:
Explaining Explanations: An Overview of Interpretability of Machine Learning,
in: 2018 IEEE 5th International Conference on Data Science and Advanced
Analytics (DSAA), 1–3 October 2018, Turin, Italy, 80–89, IEEE, https://doi.org/10.1109/dsaa.2018.00018, 2018. a
Green, J. K., Konings, A. G., Alemohammad, S. H., Berry, J., Entekhabi, D.,
Kolassa, J., Lee, J.-E., and Gentine, P.: Regionally strong feedbacks between
the atmosphere and terrestrial biosphere, Nat. Geosci., 10, 410–414,
https://doi.org/10.1038/ngeo2957, 2017. a
Green, J. K., Seneviratne, S. I., Berg, A. M., Findell, K. L., Hagemann, S.,
Lawrence, D. M., and Gentine, P.: Large influence of soil moisture on
long-term terrestrial carbon uptake, Nature, 565, 476–479,
https://doi.org/10.1038/s41586-018-0848-x, 2019. a
Guillod, B. P., Orlowsky, B., Miralles, D. G., Teuling, A. J., and Seneviratne,
S. I.: Reconciling spatial and temporal soil moisture effects on afternoon
rainfall, Nat. Commun., 6, 6443, https://doi.org/10.1038/ncomms7443, 2015. a
Guo, R., Cheng, L., Li, J., Hahn, P. R., and Liu, H.: A Survey of Learning
Causality with Data, ACM Computing Surveys, 53, 1–37,
https://doi.org/10.1145/3397269, 2021. a
Ham, Y., Kim, J., and Luo, J.: Deep learning for multi-year ENSO forecasts,
Nature, 573, 568–572, https://doi.org/10.1038/s41586-019-1559-7, 2019. a
Hartick, C., Furusho-Percot, C., Goergen, K., and Kollet, S.: An Interannual
Probabilistic Assessment of Subsurface Water Storage Over Europe Using a
Fully Coupled Terrestrial Model, Water Resour. Res., 57, e2020WR027828,
https://doi.org/10.1029/2020wr027828, 2021. a
Hersbach, H., Bell, B., Berrisford, P., Biavati, G., Horányi, A., Muñoz Sabater, J., Nicolas, J., Peubey, C., Radu, R., Rozum, I., Schepers, D., Simmons, A., Soci, C., Dee, D., and Thépaut, J.-N.: ERA5 hourly data on single levels from 1940 to present, Copernicus Climate Change Service (C3S) Climate Data Store (CDS) [data set], https://doi.org/10.24381/cds.adbb2d47, 2023. a, b, c
Hesterberg, T.: What Teachers Should Know about the Bootstrap: Resampling in
the Undergraduate Statistics Curriculum,
https://arxiv.org/abs/1411.5279 (last access: 16 April 2023), 2014. a
Holgate, C. M., Dijk, A. I. J. M. V., Evans, J. P., and Pitman, A. J.: The
Importance of the One-Dimensional Assumption in Soil Moisture – Rainfall
Depth Correlation at Varying Spatial Scales, J. Geophys. Res.-Atmos., 124, 2964–2975, https://doi.org/10.1029/2018jd029762, 2019. a
Humphrey, V., Berg, A., Ciais, P., Gentine, P., Jung, M., Reichstein, M.,
Seneviratne, S. I., and Frankenberg, C.: Soil moisture–atmosphere
feedback dominates land carbon uptake variability, Nature, 592, 65–69,
https://doi.org/10.1038/s41586-021-03325-5, 2021. a
Imamovic, A., Schlemmer, L., and Schär, C.: Collective impacts of orography
and soil moisture on the soil moisture-precipitation feedback, Geophys. Res. Lett., 44, 11682–11691, https://doi.org/10.1002/2017GL075657, 2017. a
Kingma, D. P. and Ba, J.: Adam: A Method for Stochastic Optimization,
https://arxiv.org/abs/1412.6980 (last access: 16 April 2023), 2017. a
Koster, R. D.: Regions of Strong Coupling Between Soil Moisture and
Precipitation, Science, 305, 1138–1140, https://doi.org/10.1126/science.1100217, 2004. a
LeCun, Y., Bengio, Y., and Hinton, G.: Deep learning, Nature, 521, 436–444,
https://doi.org/10.1038/nature14539, 2015. a, b
Leutwyler, D., Imamovic, A., and Schär, C.: The Continental-Scale
Soil-Moisture Precipitation Feedback in Europe with Parameterized and
Explicit Convection, J. Climate, 34, 1–56,
https://doi.org/10.1175/jcli-d-20-0415.1, 2021. a
McGovern, A., Lagerquist, R., Gagne II, D. J., Jergensen, G. E., Elmore, K. L.,
Homeyer, C. R., and Smith, T.: Making the Black Box More Transparent:
Understanding the Physical Implications of Machine Learning, B.
Am. Meteorol. Soc., 100, 2175–2199,
https://doi.org/10.1175/bams-d-18-0195.1, 2019. a
Miller, J. W., Goodman, R., and Smyth, P.: On loss functions which minimize to
conditional expected values and posterior probabilities, IEEE T.
Inform. Theor., 39, 1404–1408, https://doi.org/10.1109/18.243457, 1993. a
Molnar, C.: Interpretable Machine Learning,
https://christophm.github.io/interpretable-ml-book/ (last access: 16 April 2023), 2019. a
Montavon, G., Samek, W., and Müller, K.: Methods for interpreting and
understanding deep neural networks, Digit. Signal Process., 73, 1–15,
https://doi.org/10.1016/j.dsp.2017.10.011, 2018. a
Padarian, J., McBratney, A. B., and Minasny, B.: Game theory interpretation of digital soil mapping convolutional neural networks, SOIL, 6, 389–397, https://doi.org/10.5194/soil-6-389-2020, 2020. a
Paszke, A., Gross, S., Massa, F., Lerer, A., Bradbury, J., Chanan, G., Killeen,
T., Lin, Z., Gimelshein, N., Antiga, L., Desmaison, A., Kopf, A., Yang, E.,
DeVito, Z., Raison, M., Tejani, A., Chilamkurthy, S., Steiner, B., Fang, L.,
Bai, J., and Chintala, S.: PyTorch: An Imperative Style, High-Performance
Deep Learning Library, in: Advances in Neural Information Processing Systems
32, edited by: Wallach, H., Larochelle, H., Beygelzimer, A., d'Alché-Buc, F., Fox, E., and Garnett, R., Curran
Associates, Inc., 8026–8037,
http://papers.nips.cc/paper/9015-pytorch-an-imperative-style-high-performance-deep-learning-library.pdf (last access: 16 April 2023),
2019. a
Pearl, J.: Causal inference in statistics: An overview, Statistics Surveys, 3, 96–146,
https://doi.org/10.1214/09-ss057, 2009. a, b, c
Peters, J., Bühlmann, P., and Meinshausen, N.: Causal inference by using
invariant prediction: identification and confidence intervals, J. R. Stat.
Soc.: Series B, 78, 947–1012,
https://doi.org/10.1111/rssb.12167, 2016. a
Reichstein, M., Camps-Valls, G., Stevens, B., Jung, M., Denzler, J.,
Carvalhais, N., and Prabhat: Deep learning and process understanding for
data-driven Earth system science, Nature, 566, 195–204,
https://doi.org/10.1038/s41586-019-0912-1, 2019. a
Ronneberger, O., Fischer, P., and Brox, T.: U-Net: Convolutional Networks for
Biomedical Image Segmentation, in: Medical Image Computing and
Computer-Assisted Intervention – MICCAI 2015, edited by: Navab, N.,
Hornegger, J., Wells, W. M., and Frangi, A. F., Springer
International Publishing, Cham, 234–241,
https://arxiv.org/abs/1505.04597 (last access: 16 April 2023), 2015. a, b
Roscher, R., Bohn, B., Duarte, M. F., and Garcke, J.: Explainable Machine
Learning for Scientific Insights and Discoveries, IEEE Access, 8,
42200–42216, https://doi.org/10.1109/ACCESS.2020.2976199, 2020. a
Runge, J.: Causal network reconstruction from time series: From theoretical
assumptions to practical estimation, Chaos, 28, 075310, https://doi.org/10.1063/1.5025050, 2018. a
Runge, J., Bathiany, S., Bollt, E., Camps-Valls, G., Coumou, D., Deyle, E.,
Glymour, C., Kretschmer, M., Mahecha, M. D., Muñoz-Marí, J., van
Nes, E. H., Peters, J., Quax, R., Reichstein, M., Scheffer, M.,
Schölkopf, B., Spirtes, P., Sugihara, G., Sun, J., Zhang, K., and
Zscheischler, J.: Inferring causation from time series in Earth system
sciences, Nat. Commun., 10, 2553, https://doi.org/10.1038/s41467-019-10105-3, 2019. a
Samek, W., Montavon, G., Lapuschkin, S., Anders, C. J., and Müller,
K. R.: Explaining Deep Neural Networks and Beyond: A Review of Methods and
Applications, Proc. IEEE, 109, 247–278,
https://doi.org/10.1109/JPROC.2021.3060483, 2021. a
Santanello, J. A., Dirmeyer, P. A., Ferguson, C. R., Findell, K. L., Tawfik,
A. B., Berg, A., Ek, M., Gentine, P., Guillod, B. P., van Heerwaarden, C.,
Roundy, J., and Wulfmeyer, V.: Land–Atmosphere Interactions: The
LoCo Perspective, B. Am. Meteorol. Soc., 99,
1253–1272, https://doi.org/10.1175/bams-d-17-0001.1, 2018. a, b
Schumacher, D. L., Keune, J., van Heerwaarden, C. C., de Arellano, J. V.-G.,
Teuling, A. J., and Miralles, D. G.: Amplification of mega-heatwaves through
heat torrents fuelled by upwind drought, Nat. Geosci., 12, 712–717,
https://doi.org/10.1038/s41561-019-0431-6, 2019. a
Schwingshackl, C., Hirschi, M., and Seneviratne, S. I.: Quantifying
Spatiotemporal Variations of Soil Moisture Control on Surface Energy Balance
and Near-Surface Air Temperature, J. Climate, 30, 7105–7124,
https://doi.org/10.1175/jcli-d-16-0727.1, 2017. a
Seneviratne, S. I., Lüthi, D., Litschi, M., and Schär, C.:
Land–atmosphere coupling and climate change in Europe, Nature,
443, 205–209, https://doi.org/10.1038/nature05095, 2006. a, b
Seneviratne, S. I., Corti, T., Davin, E. L., Hirschi, M., Jaeger, E. B.,
Lehner, I., Orlowsky, B., and Teuling, A. J.: Investigating soil
moisture–climate interactions in a changing climate: A review,
Earth-Sci. Rev., 99, 125–161, https://doi.org/10.1016/j.earscirev.2010.02.004,
2010. a, b, c
Taylor, C. M.: Detecting soil moisture impacts on convective initiation in
Europe, Geophys. Res. Lett., 42, 4631–4638,
https://doi.org/10.1002/2015gl064030, 2015. a, b
Taylor, C. M., Gounou, A., Guichard, F., Harris, P. P., Ellis, R. J., Couvreux,
F., and Kauwe, M. D.: Frequency of Sahelian storm initiation enhanced over
mesoscale soil-moisture patterns, Nat. Geosci., 4, 430–433,
https://doi.org/10.1038/ngeo1173, 2011. a
Tesch, T., Kollet, S., and Garcke, J.: Variant Approach for Identifying
Spurious Relations That Deep Learning Models Learn, Front. Water, 3,
114, https://doi.org/10.3389/frwa.2021.745563, 2021. a
Tesch, T., Kollet, S., and Garcke, J.: Causal deep learning models for studying the Earth system: soil moisture-precipitation coupling in ERA5 data across Europe – Software Code, Zenodo [code], https://doi.org/10.5281/zenodo.6385040, 2022.
a
Tietz, M., Fan, T. J., Nouri, D., Bossan, B., and skorch Developers: skorch:
A scikit-learn compatible neural network library that wraps PyTorch,
https://skorch.readthedocs.io/en/stable/ (last access: 16 April 2023), 2017. a
Toms, B. A., Barnes, E. A., and Ebert-Uphoff, I.: Physically Interpretable
Neural Networks for the Geosciences: Applications to Earth System
Variability, J. Adv. Model. Earth Sy., 12,
e2019MS002002, https://doi.org/10.1029/2019ms002002, 2020. a
Tuttle, S. and Salvucci, G.: Empirical evidence of contrasting soil
moisture–precipitation feedbacks across the United States,
Science, 352, 825–828, https://doi.org/10.1126/science.aaa7185, 2016. a, b
Tuttle, S. E. and Salvucci, G. D.: Confounding factors in determining causal
soil moisture-precipitation feedback, Water Resour. Res., 53,
5531–5544, https://doi.org/10.1002/2016wr019869, 2017. a
Welty, J. and Zeng, X.: Does Soil Moisture Affect Warm Season Precipitation
Over the Southern Great Plains?, Geophys. Res. Lett., 45,
7866–7873, https://doi.org/10.1029/2018gl078598, 2018. a
Witte, J., Henckel, L., Maathuis, M. H., and Didelez, V.: On Efficient
Adjustment in Causal Graphs, J. Mach. Learn. Res., 21, 1–45,
https://doi.org/10.48550/arXiv.2002.06825, 2020. a
Zhang, Q. and Zhu, S.: Visual interpretability for deep learning: a survey,
Frontiers Inf. Technol. Electronic Eng., 19, 27–39,
https://doi.org/10.1631/fitee.1700808, 2018. a
Executive editor
Many papers are currently being published applying deep learning to geoscientific applications. However, most of them only offer proof of concept results on highly idealised scenarios. This paper combines deep learning approaches with the structural causal models popularized by the work of Judea Pearl, and it applies this methodology to a real problem, analyzing soil moisture-precipitation coupling in climate reanalysis data. In contrast to many papers in this field, this promises actual insight in the scientific application of the work.
Many papers are currently being published applying deep learning to geoscientific applications....
Short summary
A recent statistical approach for studying relations in the Earth system is to train deep learning (DL) models to predict Earth system variables given one or several others and use interpretable DL to analyze the relations learned by the models. Here, we propose to combine the approach with a theorem from causality research to ensure that the deep learning model learns causal rather than spurious relations. As an example, we apply the method to study soil-moisture–precipitation coupling.
A recent statistical approach for studying relations in the Earth system is to train deep...