Artificial intelligence reconstructs missing climate information - newest developments

<p>Historical temperature measurements are the basis of global climate datasets like HadCRUT4. This dataset contains many missing values, particularly for periods before the mid-twentieth century, although recent years are also incomplete. Here we demonstrate that artificial intelligence can skilfully fill these observational gaps when combined with numerical climate model data. We show that recently developed image inpainting techniques perform accurate monthly reconstructions via transfer learning using either 20CR (Twentieth-Century Reanalysis) or the CMIP5 (Coupled Model Intercomparison Project Phase 5) experiments. The resulting global annual mean temperature time series exhibit high Pearson correlation coefficients (≥0.9941) and low root mean squared errors (≤0.0547 °C) as compared with the original data. These techniques also provide advantages relative to state-of-the-art kriging interpolation and principal component analysis-based infilling. When applied to HadCRUT4, our method restores a missing spatial pattern of the documented El Niño from July 1877. With respect to the global mean temperature time series, a HadCRUT4 reconstruction by our method points to a cooler nineteenth century, a less apparent hiatus in the twenty-first century, an even warmer 2016 being the warmest year on record and a stronger global trend between 1850 and 2018 relative to previous estimates. We propose image inpainting as an approach to reconstruct missing climate information and thereby reduce uncertainties and biases in climate records.</p> <p>As published in:</p> <p>Kadow, C., Hall, D.M. & Ulbrich, U. Artificial intelligence reconstructs missing climate information. <em>Nat. Geosci.</em> <strong>13, </strong>408–413 (2020). https://doi.org/10.1038/s41561-020-0582-5</p> <p>Newest developments around the technology will be presented.</p> <p> </p>

Estimation of depth-resolved profiles of soil thermal diffusivity from temperature time series and uncertainty quantification

Improving the quantification of soil thermal and physical properties is key to achieving a better understanding and prediction of soil hydro-biogeochemical processes and their responses to changes in atmospheric forcing. Obtaining such information at numerous locations and/or over time with conventional soil sampling is challenging. The increasing availability of low-cost, vertically resolved temperature sensor arrays offers promise for improving the estimation of soil thermal properties from temperature time series, and the possible indirect estimation of physical properties. Still, the reliability and limitations of such an approach needs to be assessed. In the present study, we develop a parameter estimation approach based on a combination of thermal modeling, sliding time-windows, Bayesian inference, and Markov chain Monte Carlo simulation to estimate thermal diffusivity and its uncertainty over time, at numerous locations and at an unprecedented vertical spatial resolution (i.e., down to 5 to 10 cm vertical resolution) from soil temperature time series. We provide the necessary framework to assess under which environmental conditions (soil temperature gradient, fluctuations, and trend), temperature sensor characteristics (bias and level of noise) and deployment geometries (sensor number and position) soil thermal diffusivity can be reliably inferred. We validate the method with synthetic experiments and field studies. The synthetic experiments show that in the presence of median diurnal fluctuations ≥ 1.5 °C at 5 cm below the ground surface, temperature gradients > 2 °C m−1, and a sliding time-window of at least 4 days, the proposed method provides reliable depth-resolved thermal diffusivity estimates with percentage errors ≤ 10 % and posterior relative standard deviations ≤ 5 % up to 1 m depth. Reliable thermal diffusivity under such environmental conditions also requires temperature sensors spaced precisely (with few-millimeter accuracy), with a level of noise ≤ 0.02 °C, and with a bias defined by a standard deviation ≤ 0.01 °C. Finally, the application of the developed approach to field data indicates significant repeatability in results and similarity with independent measurements, as well as promise in using a sliding time-window to estimate temporal changes in soil thermal diffusivity, as needed to potentially capture changes in carbon or water content.

Spatial-temporal analysis of temperature in Lake Shira based on long-term observations

Long-term in-situ measurements of temperature were conducted in lake Shira during 2013-2015. The principal component analysis of temperature time series allowed to identify period of generation and propagation of internal waves. The spectral analysis revealed the dominance of the oscillations with periods of 21.3, 10.6 and 5.3 h.

Neural Approach in Short-Term Outdoor Temperature Prediction for Application in HVAC Systems

An accurate air-temperature prediction can provide the energy consumption and system load in advance, both of which are crucial in HVAC (heating, ventilation, air conditioning) system operation optimisation as a way of reducing energy losses, operating costs, as well as pollution and dust emissions while maintaining residents’ thermal comfort. This article presents the results of an outdoor air-temperature time-series prediction for a multifamily building with the use of artificial neural networks during the heating period (October–May). The aim of the research was to analyse in detail the created neural models with a view to select the best combination of predictors and the optimal number of neurons in a hidden layer. To meet that task, the Akaike information criterion was used. The most accurate results were obtained by MLP 3-3-1 (r = 0.986, AIC = 1300.098, SSE = 4467.109), with the ambient-air-temperature time series observed 1, 2, and 24 h before the prognostic temperature as predictors. The AIC proved to be a useful method for the optimum model selection in a machine-learning modelling. What is more, neural network models provide the most accurate prediction, when compared with LR and SVR. Additionally, the obtained temperature predictions were used in HVAC applications: entering-water temperature and indoor temperature modelling.

Updated Moving Forecasting Model of Air Maximum Temperature

In the current study, a moving forecasting model is used for the purpose of forecasting maximum air temperature. A number of recordings are used for building the AR model and next, to forecasting some temperature values ahead. Then the AR model coefficients are updating due to shifting the training sample by adding new temperature values in order to involve the change in temperature time series behaviour. The current work shows a high performance all over the temperature time series, which considered in the analysis.

Can Simple Machine Learning Tools Extend and Improve Temperature-Based Methods to Infer Streambed Flux?

Temperature-based methods have been developed to infer 1D vertical exchange flux between a stream and the subsurface. Current analyses rely on fitting physically based analytical and numerical models to temperature time series measured at multiple depths to infer daily average flux. These methods have seen wide use in hydrologic science despite strong simplifying assumptions including a lack of consideration of model structural error or the impacts of multidimensional flow or the impacts of transient streambed hydraulic properties. We performed a “perfect-model experiment” investigation to examine whether regression trees, with and without gradient boosting, can extract sufficient information from model-generated subsurface temperature time series, with and without added measurement error, to infer the corresponding exchange flux time series at the streambed surface. Using model-generated, synthetic data allowed us to assess the basic limitations to the use of machine learning; further examination of real data is only warranted if the method can be shown to perform well under these ideal conditions. We also examined whether the inherent feature importance analyses of tree-based machine learning methods can be used to optimize monitoring networks for exchange flux inference.