Advancing global sea ice prediction capabilities using a fully coupled climate model with integrated machine learning

Global, Arctic, and Antarctic sea ice volume predictions using score-driven threshold climate models

<strong class="journal-contentHeaderColor">Abstract.</strong> Over the past three decades, discordant trends in sea ice extent have been observed between the Arctic and Antarctic regions. Arctic sea ice extent has been characterised by a rapid decline, whereas Antarctic sea ice extent, while highly variable inter-annually, has tended to increase. Climate models have so far failed to capture these trends. Coupled with the limited pre-1970 sea ice dataset, this poses a significant challenge to quantifying the mechanisms responsible for driving such trends. However, historical records from early Antarctic expeditions contain a wealth of information regarding the nature and concentration of sea ice. Such records have been under-utilised, and their analysis may enhance our understanding of recent Antarctic sea ice variability. For the purpose of this study, 9 records from 8 Antarctic expeditions have been examined. Summer sea ice positions recorded during 1820&ndash;1843 have been compared to satellite observations from 1987&ndash;2017, as well as historical data for the period 1897&ndash;1917. Through analysis of these three time series, estimations for summer sea ice extent in the Weddell Sea, during the early nineteenth century have been produced. The key findings of this study indicate a nineteenth century average core summer northernmost sea ice latitude in much of the Weddell Sea that was further north than during the modern era, with nineteenth century February having significantly more sea ice by all measures. However, late summer sea ice was most extensive in the early years of the twentieth century.

<strong class="journal-contentHeaderColor">Abstract.</strong> Over the past three decades, discordant trends in sea ice extent have been observed between the Arctic and Antarctic regions. Arctic sea ice extent has been characterised by a rapid decline, whereas Antarctic sea ice extent, while highly variable inter-annually, has tended to increase. Climate models have so far failed to capture these trends. Coupled with the limited pre-1970 sea ice dataset, this poses a significant challenge to quantifying the mechanisms responsible for driving such trends. However, historical records from early Antarctic expeditions contain a wealth of information regarding the nature and concentration of sea ice. Such records have been under-utilised, and their analysis may enhance our understanding of recent Antarctic sea ice variability. For the purpose of this study, 9 records from 8 Antarctic expeditions have been examined. Summer sea ice positions recorded during 1820&ndash;1843 have been compared to satellite observations from 1987&ndash;2017, as well as historical data for the period 1897&ndash;1917. Through analysis of these three time series, estimations for summer sea ice extent in the Weddell Sea, during the early nineteenth century have been produced. The key findings of this study indicate a nineteenth century average core summer northernmost sea ice latitude in much of the Weddell Sea that was further north than during the modern era, with nineteenth century February having significantly more sea ice by all measures. However, late summer sea ice was most extensive in the early years of the twentieth century.

The Antarctic sea ice extent slowly expanded through the four-decade-long satellite era until 2014 when the expansion came to a halt, followed by a rapid contraction in the next couple of years. This sudden unexpected trend reversal has sparked considerable research interest and several mechanisms have been proposed to explain it; however, much remains to be explored. In this study, we show that the long-term increasing trend in the Antarctic sea ice extent and its recent reversal can be largely explained by the first, second and fourth empirical orthogonal function mode of sea ice variability in austral summer, autumn and spring, respectively. We illustrate that the sea ice variability represented by the three modes is mostly consistent with what is expected from the anomalous atmospheric circulations associated with planetary wavetrains that are triggered by anomalous sea surface temperature (SST) and convective activities over the Southern Indian and Pacific Oceans. More specifically, the results suggest a teleconnection between the increasing periods in the Antarctic sea ice extent in the past four decades and the positive SST anomalies over the southeastern Indian Ocean and the western tropical Pacific Ocean. The opposite occurs over the decreasing period. Accordingly, the same mechanisms, in different phases, have been associated with the periods of increasing and decreasing Antarctic sea ice extent.

Integrating machine learning (ML) models into healthcare systems is a rapidly evolving field with the potential to revolutionize care delivery. This study aimed to classify fertility rates and identify significant predictors using ML models among reproductive women in Ethiopia. This study utilized eight ML models in 5864 reproductive-age women using Ethiopian Demographic Health Survey (EDHS), 2019 data. Phyton programming language was used to develop these models. Predictors of fertility rate were determined using the feature important techniques. The performance of models was evaluated using accuracy, area under the curve (AUC), precision, recall, F1-score, specificity, and sensitivity. The mean age of participants was 32.7 (± 5.6) years. The random forest classifier (accuracy = 0.901 and AUC = 0.961) followed by a one-dimensional convolutional neural network (accuracy = 0.899 and AUC = 0.958), logistic regression (accuracy = 0.874 and AUC = 0.937), and gradient boost classifier (accuracy = 0.851 and AUC 0.927) were the top performing ML models. Family size, age, occupation, and education with an average importance score of 0.198, 0.151, 0.118, and 0.081, respectively were the top significant predictors of the fertility rate. The best ML models to classify and predict fertility rates were random forest, one-dimensional convolutional neural network, logistic regression, and gradient boost classifier. The findings on important factors of fertility rate can inform targeted public health, programs that address disparities related to family size, occupation, education, and other socioeconomic factors.

In stark contrast to the Arctic, there have been statistically significant positive trends in total Antarctic sea ice extent since 1979. However, the short and highly variable nature of observed Antarctic sea ice extent limits the ability to fully understand the historical context of these recent changes. To meet this challenge, we have created robust, observation-based reconstruction ensembles of seasonal Antarctic sea ice extent since 1905. Using these reconstructions, here we show that the observed period since 1979 is the only time all four seasons demonstrate significant increases in total Antarctic sea ice in the context of the twentieth century and that the observed increases are juxtaposed against statistically significant decreases throughout much of the early and middle twentieth century. These reconstructions provide reliable estimates of seasonally resolved total Antarctic sea ice extent and are skilful enough to better understand aspects of air–sea–ice interactions within the Antarctic climate system. Antarctic sea ice extent is thought to be stable or increasing, in contrast to Arctic declines. Estimates of seasonal Antarctic sea ice from reconstructions show that increases are confined to the satellite era, post-1979, with substantial decreases in the early and mid-twentieth century.

Changes in deep-ocean circulation and stratification have been argued to contribute to climatic shifts between glacial and interglacial climates by affecting the atmospheric carbon dioxide concentrations. It has been recently proposed that such changes are associated with variations in Antarctic sea ice through two possible mechanisms: an increased latitudinal extent of Antarctic sea ice and an increased rate of Antarctic sea ice formation. Both mechanisms lead to an upward shift of the Atlantic meridional overturning circulation (AMOC) above depths where diapycnal mixing is strong (above 2000 m), thus decoupling the AMOC from the abyssal overturning circulation. Here, these two hypotheses are tested using a series of idealized two-basin ocean simulations. To investigate independently the effect of an increased latitudinal ice extent from the effect of an increased ice formation rate, sea ice is parameterized as a latitude strip over which the buoyancy flux is negative. The results suggest that both mechanisms can effectively decouple the two cells of the meridional overturning circulation (MOC), and that their effects are additive. To illustrate the role of Antarctic sea ice in decoupling the AMOC and the abyssal overturning cell, the age of deep-water masses is estimated. An increase in both the sea ice extent and its formation rate yields a dramatic “aging” of deep-water masses if the sea ice is thick and acts as a lid, suppressing air–sea fluxes. The key role of vertical mixing is highlighted by comparing results using different profiles of vertical diffusivity. The implications of an increase in water mass ages for storing carbon in the deep ocean are discussed.

Antarctic sea ice exerts great influence on Earth’s climate by controlling the exchange of heat, momentum, freshwater, and gases between the atmosphere and ocean. Antarctic sea ice extent has undergone a multidecadal slight increase followed by a substantial decline since 2016. Here we utilize a 300-yr sea ice data assimilation reconstruction and two NOAA/GFDL and five CMIP6 model simulations to demonstrate a multidecadal variability of Antarctic sea ice extent. Stronger westerlies associated with the Southern Annular Mode (SAM) enhance the upwelling of warm and saline water from the subsurface ocean. The consequent salinity increase weakens the upper-ocean stratification, induces deep convection, and in turn brings more subsurface warm and saline water to the surface. This salinity-convection feedback triggered by the SAM provides favorable conditions for multidecadal sea ice decrease. Processes acting in reverse are found to cause sea ice increase, although it evolves slower than sea ice decrease.

Machine learning (ML) models provide more choices to patients with diabetes mellitus (DM) to more properly manage blood glucose (BG) levels. However, because of numerous types of ML algorithms, choosing an appropriate model is vitally important. In a systematic review and network meta-analysis, this study aimed to comprehensively assess the performance of ML models in predicting BG levels. In addition, we assessed ML models used to detect and predict adverse BG (hypoglycemia) events by calculating pooled estimates of sensitivity and specificity. PubMed, Embase, Web of Science, and Institute of Electrical and Electronics Engineers Explore databases were systematically searched for studies on predicting BG levels and predicting or detecting adverse BG events using ML models, from inception to November 2022. Studies that assessed the performance of different ML models in predicting or detecting BG levels or adverse BG events of patients with DM were included. Studies with no derivation or performance metrics of ML models were excluded. The Quality Assessment of Diagnostic Accuracy Studies tool was applied to assess the quality of included studies. Primary outcomes were the relative ranking of ML models for predicting BG levels in different prediction horizons (PHs) and pooled estimates of the sensitivity and specificity of ML models in detecting or predicting adverse BG events. In total, 46 eligible studies were included for meta-analysis. Regarding ML models for predicting BG levels, the means of the absolute root mean square error (RMSE) in a PH of 15, 30, 45, and 60 minutes were 18.88 (SD 19.71), 21.40 (SD 12.56), 21.27 (SD 5.17), and 30.01 (SD 7.23) mg/dL, respectively. The neural network model (NNM) showed the highest relative performance in different PHs. Furthermore, the pooled estimates of the positive likelihood ratio and the negative likelihood ratio of ML models were 8.3 (95% CI 5.7-12.0) and 0.31 (95% CI 0.22-0.44), respectively, for predicting hypoglycemia and 2.4 (95% CI 1.6-3.7) and 0.37 (95% CI 0.29-0.46), respectively, for detecting hypoglycemia. Statistically significant high heterogeneity was detected in all subgroups, with different sources of heterogeneity. For predicting precise BG levels, the RMSE increases with a rise in the PH, and the NNM shows the highest relative performance among all the ML models. Meanwhile, current ML models have sufficient ability to predict adverse BG events, while their ability to detect adverse BG events needs to be enhanced. PROSPERO CRD42022375250; https://www.crd.york.ac.uk/prospero/display_record.php?RecordID=375250.

Geophysical constraints on the Antarctic sea ice cover

Antarctic sea ice cover experienced an abrupt decline in 2016, transitioning from a record maximum state to a record minimum state. However, the drivers of this rapid retreat are currently not well understood. Therefore, it is difficult to determine whether this signals the start of a long term melting trend, as has been long anticipated by climate models, or is an isolated episode of internal climate variability. In this study, we utilise the CMIP6 archive to understand if internal climate variability could be responsible for this Antarctic sea ice anomaly, and if so what the primary atmospheric and oceanic drivers are. This involves examining composites of the tropical teleconnections, subsurface ocean heat content, and high latitude atmospheric variability preceding extreme Antarctic sea ice anomalies in CMIP6 simulations. The primary objective is to elucidate the multifaceted factors influencing these extreme events, specifically addressing the 2016-2017 sea ice retreat, with lessons for 2023&amp;#8217;s extreme Antarctic sea ice state. Initial results indicate that such events are possible in the absence of anthropogenic emissions in some climate models, although the occurrences are considered rare. We also show that that using the limited observed record alone will underestimate the interannual variability of the Antarctic sea ice cover and therefore overestimate how rare such an anomaly would be. In fact, if we extend the observed record further back using statistical reconstructions, rapid declines of sea ice extent occurred in the early and mid 20th century. &amp;#160;Our results highlight the importance of internal climate variability in the Southern high latitudes and advance our understanding of the drivers and predictability of Antarctic sea ice changes. We discuss the implications of this work for 2023&amp;#8217;s record Antarctic sea ice anomaly.

Systematic literature review of machine learning based software development effort estimation models

This study focuses on using various machine learning (ML) models to evaluate the shear behaviors of ultra-high-performance concrete (UHPC) beams reinforced with glass fiber-reinforced polymer (GFRP) bars. The main objective of the study is to predict the shear strength of UHPC beams reinforced with GFRP bars using ML models. We use four different ML models: support vector machine (SVM), artificial neural network (ANN), random forest (R.F.), and extreme gradient boosting (XGBoost). The experimental database used in the study is acquired from various literature sources and comprises 54 test observations with 11 input features. These input features are likely parameters related to the composition, geometry, and properties of the UHPC beams and GFRP bars. To ensure the ML models' generalizability and scalability, random search methods are utilized to tune the hyperparameters of the algorithms. This tuning process helps improve the performance of the models when predicting the shear strength. The study uses the ACI318M-14 and Eurocode 2 standard building codes to predict the shear capacity behavior of GFRP bars-reinforced UHPC I-shaped beams. The ML models' predictions are compared to the results obtained from these building code standards. According to the findings, the XGBoost model demonstrates the highest predictive test performance among the investigated ML models. The study employs the SHAP (SHapley Additive exPlanations) analysis to assess the significance of each input parameter in the ML models' predictive capabilities. A Taylor diagram is used to statistically compare the accuracy of the ML models. This study concludes that ML models, particularly XGBoost, can effectively predict the shear capacity behavior of GFRP bars-reinforced UHPC I-shaped beams.

The medical imaging community has embraced Machine Learning (ML) as evidenced by the rapid increase in the number of ML models being developed, but validating and deploying these models in the clinic remains a challenge. The engineering involved in integrating and assessing the efficacy of ML models within the clinical workflow is complex. This paper presents a general-purpose, end-to-end, clinically integrated ML model deployment and validation system implemented at UCSF. Engineering and usability challenges and results from 3 use cases are presented. A generalized validation system based on free, open-source software (OSS) was implemented, connecting clinical imaging modalities, the Picture Archiving and Communication System (PACS), and an ML inference server. ML pipelines were implemented in NVIDIA’s Clara Deploy framework with results and clinician feedback stored in a customized XNAT instance, separate from the clinical record but linked from within PACS. Prospective clinical validation studies of 3 ML models were conducted, with data routed from multiple clinical imaging modalities and PACS. Completed validation studies provided expert clinical feedback on model performance and usability, plus system reliability and performance metrics. Clinical validation of ML models entails assessing model performance, impact on clinical infrastructure, robustness, and usability. Study results must be easily accessible to participating clinicians but remain outside the clinical record. Building a system that generalizes and scales across multiple ML models takes the concerted effort of software engineers, clinicians, data scientists, and system administrators, and benefits from the use of modular OSS. The present work provides a template for institutions looking to translate and clinically validate ML models in the clinic, together with required resources and expected challenges.

In the Antarctic, sea ice algae use the highly dynamic sea ice as a platform for growth. Antarctic sea ice extent has recently been highly variable, first showing a slight increase and then a record decrease starting in 2016. We investigated the response of Antarctic ice algal habitat to variations in sea ice and other environmental forcings during 2004–2019. Combining an ice growth model, remote sensing and reanalysis data, and a radiative transfer model, we assessed whether light penetration to the bottom ice was sufficient for ice algal growth. Trends in the inputs over the 16 years were relatively small: there were no changes in ice thickness or bottom ice melt date, a 6.4% decrease in snow depth, a 1.2% decrease in incident light, and a 0.8°C decrease in air temperatures. Eighty‐one percent of the sea ice cover was habitable by ice algae for ≥14 days each year. The Antarctic has a larger extent and duration of potential ice algal habitat than the Arctic. Over time, the spatially averaged seasonal duration of habitat increased because a higher proportion of each pixel became habitable on average, compensating for the 2016–2019 reduction in sea ice extent. The spatial variability in potential habitat was strikingly high, even within geographic sectors. Bottom ice melt date (bloom termination) far surpassed other environmental factors in explaining variation (45%) in ice algal habitat on the 25 km scale. Because melt date depends on the ice‐atmosphere heat balance, Antarctic ice algal habitat may be highly sensitive to future climate changes.