Tropospheric ozone (O3) is a greenhouse gas which can absorb heat and make the weather even hotter during extreme heatwaves. Besides, it is an influential ground-level air pollutant which can severely damage the environment. Thus evaluating the importance of various factors related to the O3 formation process is essential. However, O3 simulated by the available climate models exhibits large variance in different places, indicating the insufficiency of models in explaining the O3 formation process correctly. In this paper, we aim to identify and understand the impact of various factors on O3 formation and predict the O3 concentrations under different pollution-reduced and climate change scenarios. We employ six supervised methods to estimate the observed O3 using fourteen meteorological and chemical variables. We find that the deep neural network (DNN) and long short-term memory (LSTM) based models can predict O3 concentrations accurately. We also demonstrate the importance of several variables in this prediction task. The results suggest that while Nitrogen Oxides negatively contributes to predicting O3, solar radiation makes a significantly positive contribution. Furthermore, we apply our two best models on O3 prediction under different global warming and pollution reduction scenarios to improve the policy-making decisions in the O3 reduction.

To improve the physical understanding and the predictions of complex dynamic systems, such as ocean dynamics and weather predictions, it is of paramount interest to identify interpretable models from coarsely and off-grid sampled observations. In this work we investigate how deep learning can improve model discovery of partial differential equations when the spacing between sensors is large and the samples are not placed on a grid. We show how leveraging physics informed neural network interpolation and automatic differentiation, allow to better fit the data and its spatiotemporal derivatives, compared to more classic spline interpolation and numerical differentiation techniques. As a result, deep learning based model discovery allows to recover the underlying equations, even when sensors are placed further apart than the data’s characteristic length scale and in the presence of high noise levels. We illustrate our claims on both synthetic and experimental data sets where combinations of physical processes such as (non)-linear advection, reaction and diffusion are correctly identified.

As climate change increases the intensity of natural disasters, society needs better tools for adaptation. Floods, for example, are the most frequent natural disaster, and better tools for flood risk communication could increase the support for flood-resilient infrastructure development. Our work aims to enable more visual communication of large-scale climate impacts via visualizing the output of coastal flood models as satellite imagery. We propose the first deep learning pipeline to ensure physical-consistency in synthetic visual satellite imagery. We advanced a state-of-the-art GAN called pix2pixHD, such that it produces imagery that is physically-consistent with the output of an expert-validated storm surge model (NOAA SLOSH). By evaluating the imagery relative to physics-based flood maps, we find that our proposed framework outperforms baseline models in both physical-consistency and photorealism. We envision our work to be the first step towards a global visualization of how climate change shapes our landscape. Continuing on this path, we show that the proposed pipeline generalizes to visualize arctic sea ice melt. We also publish a dataset of over 25k labelled image-pairs to study image-to-image translation in Earth observation.

Short papers: - PCE-PINNs: Physics-Informed Neural Networks for Uncertainty Propagation in Ocean Modeling. Björn Lütjens (MIT), Mark Veillette (MIT Lincoln Laboratory), Dava Newman (MIT), and Cait Crawford (IBM). abstract Generative modeling of spatio-temporal weather patterns with extreme event conditioning. Konstantin Klemmer (University of Warwick), Sudipan Saha (Technical University of Munich), Matthias Kahl (Technical University of Munich), Tianlin Xu (London School of Economics and Political Science), and Xiaoxiang Zhu (Technical University of Munich). - CropGym: A reinforcement learning environment for crop management. Hiske Overweg, Herman Berghuijs, and Ioannis N. Athanasiadis (Wageningen University and Research). abstract - Ontology-based Knowledge Modelling for Ocean and Climate Change Applications. Ana C. Muñoz, Luis Martí, and Nayat Sanchez-Pi (Inria Chile Research Center). abstract - Frost Forecasting Model using Graph Neural Networks with Spatio-Temporal Attention Hernán Lira, Luis Martí, and Nayat Sanchez-Pi (Inria Chile Research Center). abstract

We present a study using a class of post-hoc local explanation methods i.e., feature importance methods for “understanding” a deep learning (DL) emulator of climate. Specifically, we consider a multiple-input-single-output emulator that uses a DenseNet encoder-decoder architecture and is trained to predict interannual variations of sea surface temperature (SST) at 1, 6, and 9 month lead times using the preceding 36 months of (appropriately filtered) SST data. First, feature importance methods are employed for individual predictions to spatio-temporally identify input features that are important for model prediction at chosen geographical regions and chosen prediction lead times. In a second step, we also examine the behavior of feature importance in a generalized sense by considering an aggregation of the importance heatmaps over training samples. We find that: 1) the climate emulator’s prediction at any geographical location depends dominantly on a small neighborhood around it; 2) the longer the prediction lead time, the further back the “importance” extends; and 3) to leading order, the temporal decay of“importance” is independent of geographical location. An ablation experiment is adopted to verify the findings. From the perspective of climate dynamics, these findings suggest a dominant role for local processes and a negligible role for re-mote teleconnections at the spatial and temporal scales we consider. From the perspective of network architecture, the spatio-temporal relations between the inputs and outputs we find suggest potential model refinements. We discuss further extensions of our methods, some of which we are considering in ongoing work.

The carbon pump of the world’s oceans plays a vital role in the biosphere and climate of the earth, urging improved understanding of the functions and influences of the oceans for climate change analyses. State-of-the-art techniques are required to develop models that can capture the complexity of ocean currents and temperature flows. We will explore the benefits of using physics informed neural networks (PINNs) for solving partial differential equations related to ocean modeling, such as the wave, shallow water, and advection-diffusion equations. PINNs account for adherence to physical laws in order to improve learning and generalization. However, in this work, we show that we observe worse training and generalizability results, contrary to recent publications.

With the decrease of hardware costs, stationary hydrophones are increasingly deployed in the marine environment to record animal vocalizations amidst ocean noise over an extended period of time. Bioacoustic data collected in this way is an important and practical source to study vocally active marine species and can make an important contribution to ecosystem monitoring. However, a main challenge of this data is the lack of annotation which many supervised neural network models rely on to learn to distinguish between noise and marine animal vocalizations. In this paper, we posit an unsupervised deep embedded clustering based on LSTM autoencoders, that aims to learn the representation of the input audio by minimizing the reconstruction loss and to simultaneously minimize a clustering loss through Kullback–Leibler divergence.

Open discussion with all participants and invited researchers.