-
Are Deep Learning Methods Suitable for Downscaling Global Climate Projections? Review and Intercomparison of Existing Models
Authors:
Jose González-Abad,
José Manuel Gutiérrez
Abstract:
Deep Learning (DL) has shown promise for downscaling global climate change projections under different approaches, including Perfect Prognosis (PP) and Regional Climate Model (RCM) emulation. Unlike emulators, PP downscaling models are trained on observational data, so it remains an open question whether they can plausibly extrapolate unseen conditions and changes in future emissions scenarios. He…
▽ More
Deep Learning (DL) has shown promise for downscaling global climate change projections under different approaches, including Perfect Prognosis (PP) and Regional Climate Model (RCM) emulation. Unlike emulators, PP downscaling models are trained on observational data, so it remains an open question whether they can plausibly extrapolate unseen conditions and changes in future emissions scenarios. Here we focus on this problem as the main drawback for the operationalization of these methods and present the results of 1) a literature review to identify state-of-the-art DL models for PP downscaling and 2) an intercomparison experiment to evaluate the performance of these models and to assess their extrapolation capability using a common experimental framework, taking into account the sensitivity of results to different training replicas. We focus on minimum and maximum temperatures and precipitation over Spain, a region with a range of climatic conditions with different influential regional processes. We conclude with a discussion of the findings, limitations of existing methods, and prospects for future development.
△ Less
Submitted 6 November, 2024;
originally announced November 2024.
-
Transferability and explainability of deep learning emulators for regional climate model projections: Perspectives for future applications
Authors:
Jorge Bano-Medina,
Maialen Iturbide,
Jesus Fernandez,
Jose Manuel Gutierrez
Abstract:
Regional climate models (RCMs) are essential tools for simulating and studying regional climate variability and change. However, their high computational cost limits the production of comprehensive ensembles of regional climate projections covering multiple scenarios and driving Global Climate Models (GCMs) across regions. RCM emulators based on deep learning models have recently been introduced a…
▽ More
Regional climate models (RCMs) are essential tools for simulating and studying regional climate variability and change. However, their high computational cost limits the production of comprehensive ensembles of regional climate projections covering multiple scenarios and driving Global Climate Models (GCMs) across regions. RCM emulators based on deep learning models have recently been introduced as a cost-effective and promising alternative that requires only short RCM simulations to train the models. Therefore, evaluating their transferability to different periods, scenarios, and GCMs becomes a pivotal and complex task in which the inherent biases of both GCMs and RCMs play a significant role. Here we focus on this problem by considering the two different emulation approaches proposed in the literature (PP and MOS, following the terminology introduced in this paper). In addition to standard evaluation techniques, we expand the analysis with methods from the field of eXplainable Artificial Intelligence (XAI), to assess the physical consistency of the empirical links learnt by the models. We find that both approaches are able to emulate certain climatological properties of RCMs for different periods and scenarios (soft transferability), but the consistency of the emulation functions differ between approaches. Whereas PP learns robust and physically meaningful patterns, MOS results are GCM-dependent and lack physical consistency in some cases. Both approaches face problems when transferring the emulation function to other GCMs, due to the existence of GCM-dependent biases (hard transferability). This limits their applicability to build ensembles of regional climate projections. We conclude by giving some prospects for future applications.
△ Less
Submitted 31 October, 2023;
originally announced November 2023.
-
Multi-variable Hard Physical Constraints for Climate Model Downscaling
Authors:
Jose González-Abad,
Álex Hernández-García,
Paula Harder,
David Rolnick,
José Manuel Gutiérrez
Abstract:
Global Climate Models (GCMs) are the primary tool to simulate climate evolution and assess the impacts of climate change. However, they often operate at a coarse spatial resolution that limits their accuracy in reproducing local-scale phenomena. Statistical downscaling methods leveraging deep learning offer a solution to this problem by approximating local-scale climate fields from coarse variable…
▽ More
Global Climate Models (GCMs) are the primary tool to simulate climate evolution and assess the impacts of climate change. However, they often operate at a coarse spatial resolution that limits their accuracy in reproducing local-scale phenomena. Statistical downscaling methods leveraging deep learning offer a solution to this problem by approximating local-scale climate fields from coarse variables, thus enabling regional GCM projections. Typically, climate fields of different variables of interest are downscaled independently, resulting in violations of fundamental physical properties across interconnected variables. This study investigates the scope of this problem and, through an application on temperature, lays the foundation for a framework introducing multi-variable hard constraints that guarantees physical relationships between groups of downscaled climate variables.
△ Less
Submitted 2 August, 2023;
originally announced August 2023.
-
Implementation of FAIR principles in the IPCC: The WGI AR6 Atlas repository
Authors:
Maialen Iturbide,
Jesús Fernández,
José M. Gutiérrez,
Anna Pirani,
David Huard,
Alaa Al Khourdajie,
Jorge Baño-Medina,
Joaquin Bedia,
Ana Casanueva,
Ezequiel Cimadevilla,
Antonio S. Cofiño,
Matteo De Felice,
Javier Diez-Sierra,
Markel García-Díez,
James Goldie,
Dimitris A. Herrera,
Sixto Herrera,
Rodrigo Manzanas,
Josipa Milovac,
Aparna Radhakrishnan,
Daniel San-Martín,
Alessandro Spinuso,
Kristen Thyng,
Claire Trenham,
Özge Yelekçi
Abstract:
The Sixth Assessment Report (AR6) of the Intergovernmental Panel on Climate Change (IPCC) has adopted the FAIR Guiding Principles. The Atlas chapter of Working Group I (WGI) is presented as a test case. Here, we describe the application of these principles in the Atlas, the challenges faced during its implementation, and those that remain for the future. We present the open source repository resul…
▽ More
The Sixth Assessment Report (AR6) of the Intergovernmental Panel on Climate Change (IPCC) has adopted the FAIR Guiding Principles. The Atlas chapter of Working Group I (WGI) is presented as a test case. Here, we describe the application of these principles in the Atlas, the challenges faced during its implementation, and those that remain for the future. We present the open source repository resulting from this process, which collects the code (including annotated Jupyter notebooks), data provenance, and some aggregated datasets underpinning the key figures in the Atlas chapter and its interactive companion (the Interactive Atlas), open to scrutiny by the scientific community and the general public. We describe the informal pilot review conducted on this repository to gather recommendations that led to significant improvements. Finally, a working example illustrates the use of the repository to produce customized regional information, extending the Interactive Atlas products and running the code interactively in a web browser using Jupyter notebooks. Atlas repository: doi:10.5281/zenodo.5171760.
△ Less
Submitted 29 April, 2022;
originally announced April 2022.
-
A Probabilistic Chemical Programmable Computer
Authors:
Abhishek Sharma,
Marcus Tze-Kiat Ng,
Juan Manuel Parrilla Gutierrez,
Yibin Jiang,
Leroy Cronin
Abstract:
The exponential growth of the power of modern digital computers is based upon the miniaturisation of vast nanoscale arrays of electronic switches, but this will be eventually constrained by fabrication limits and power dissipation. Chemical processes have the potential to scale beyond these limits performing computations through chemical reactions, yet the lack of well-defined programmability limi…
▽ More
The exponential growth of the power of modern digital computers is based upon the miniaturisation of vast nanoscale arrays of electronic switches, but this will be eventually constrained by fabrication limits and power dissipation. Chemical processes have the potential to scale beyond these limits performing computations through chemical reactions, yet the lack of well-defined programmability limits their scalability and performance. We present a hybrid digitally programmable chemical array as a probabilistic computational machine that uses chemical oscillators partitioned in interconnected cells as a computational substrate. This hybrid architecture performs efficient computation by distributing between chemical and digital domains together with error correction. The efficiency is gained by combining digital with probabilistic chemical logic based on nearest neighbour interactions and hysteresis effects. We demonstrated the implementation of one- and two- dimensional Chemical Cellular Automata and solutions to combinatorial optimization problems.
△ Less
Submitted 28 April, 2022;
originally announced April 2022.
-
The Probabilistic Backbone of Data-Driven Complex Networks: An example in Climate
Authors:
Catharina Graafland,
José M. Gutiérrez,
Juan M. López,
Diego Pazó,
Miguel A. Rodríguez
Abstract:
Correlation Networks (CNs) inherently suffer from redundant information in their network topology. Bayesian Networks (BNs), on the other hand, include only non-redundant information (from a probabilistic perspective) resulting in a sparse topology from which generalizable physical features can be extracted. We advocate the use of BNs to construct data-driven complex networks as they can be regarde…
▽ More
Correlation Networks (CNs) inherently suffer from redundant information in their network topology. Bayesian Networks (BNs), on the other hand, include only non-redundant information (from a probabilistic perspective) resulting in a sparse topology from which generalizable physical features can be extracted. We advocate the use of BNs to construct data-driven complex networks as they can be regarded as the probabilistic backbone of the underlying complex system. Results are illustrated at the hand of a global climate dataset.
△ Less
Submitted 2 November, 2020; v1 submitted 8 December, 2019;
originally announced December 2019.
-
Global Forty-Years Validation of Seasonal Precipitation Forecasts: Assessing El Niño-Driven Skill
Authors:
R. Manzanas,
M. D. Frías,
A. S. Cofiño,
J. M. Gutiérrez
Abstract:
The skill of seasonal precipitation forecasts is assessed worldwide -grid point by grid point- for the forty-years period 1961-2000. To this aim, the ENSEMBLES multi-model hindcast is considered. Although predictability varies with region, season and lead-time, results indicate that 1) significant skill is mainly located in the tropics -20 to 40% of the total land areas-, 2) overall, SON (MAM) is…
▽ More
The skill of seasonal precipitation forecasts is assessed worldwide -grid point by grid point- for the forty-years period 1961-2000. To this aim, the ENSEMBLES multi-model hindcast is considered. Although predictability varies with region, season and lead-time, results indicate that 1) significant skill is mainly located in the tropics -20 to 40% of the total land areas-, 2) overall, SON (MAM) is the most (less) skillful season and 3) predictability does not decrease noticeably from one to four months lead-time -this is so especially in northern south America and the Malay archipelago, which seem to be the most skillful regions of the world-. An analysis of teleconnections revealed that most of the skillful zones exhibit significant teleconnections with El Niño. Furthermore, models are shown to reproduce similar teleconnection patterns to those observed, especially in SON -with spatial correlations of around 0.6 in the tropics-. Moreover, these correlations are systematically higher for the skillful areas. Our results indicate that the skill found might be determined to a great extent by the models' ability to properly reproduce the observed El Niño teleconnections, i.e., the better a model simulates the El Niño teleconnections, the higher its performance is.
△ Less
Submitted 18 September, 2013;
originally announced September 2013.
