Air pollution, in particular high concentrations of particulate matter smaller than 1 µm in diameter (PM₁), continues to be a major health problem, and meteorology is known to substantially influence atmospheric PM concentrations. However, the scientific understanding of the ways in which complex interactions of meteorological factors lead to high-pollution episodes is inconclusive. In this study, a novel, data-driven approach based on empirical relationships is used to characterize and better understand the meteorology-driven component of PM₁ variability. A tree-based machine learning model is set up to reproduce concentrations of speciated PM₁ at a suburban site southwest of Paris, France, using meteorological variables as input features. The model is able to capture the majority of occurring variance of mean afternoon total PM₁ concentrations (coefficient of determination (R²) of 0.58), with model performance depending on the individual PM₁ species predicted. Based on the models, an isolation and quantification of individual, season-specific meteorological influences for process understanding at the measurement site is achieved using SHapley Additive exPlanation (SHAP) regression values. Model results suggest that winter pollution episodes are often driven by a combination of shallow mixed layer heights (MLHs), low temperatures, low wind speeds, or inflow from northeastern wind directions. Contributions of MLHs to the winter pollution episodes are quantified to be on average ∼5 µg/m³ for MLHs below <500 m a.g.l. Temperatures below freezing initiate formation processes and increase local emissions related to residential heating, amounting to a contribution to predicted PM₁ concentrations of as much as ∼9 µg/m³. Northeasterly winds are found to contribute ∼5 µg/m³ to predicted PM₁ concentrations (combined effects of u- and v-wind components), by advecting particles from source regions, e.g. central Europe or the Paris region. Meteorological drivers of unusually high PM₁ concentrations in summer are temperatures above ∼25 ^∘C (contributions of up to ∼2.5 µg/m³), dry spells of several days (maximum contributions of ∼1.5 µg/m³), and wind speeds below ∼2 m/s (maximum contributions of ∼3 µg/m³), which cause a lack of dispersion. High-resolution case studies are conducted showing a large variability of processes that can lead to high-pollution episodes. The identification of these meteorological conditions that increase air pollution could help policy makers to adapt policy measures, issue warnings to the public, or assess the effectiveness of air pollution measures.

Satellite altimeters have been used to monitor Arctic sea ice thickness since the early 2000s. In order to estimate sea ice thickness from satellite altimeter data, leads (i.e., cracks between ice floes) should first be identified for the calculation of sea ice freeboard. In this study, we proposed novel approaches for lead detection using two machine learning algorithms: decision trees and random forest. CryoSat-2 satellite data collected in March and April of 2011–2014 over the Arctic region were used to extract waveform parameters that show the characteristics of leads, ice floes and ocean, including stack standard deviation, stack skewness, stack kurtosis, pulse peakiness and backscatter sigma-0. The parameters were used to identify leads in the machine learning models. Results show that the proposed approaches, with overall accuracy >90%, produced much better performance than existing lead detection methods based on simple thresholding approaches. Sea ice thickness estimated based on the machine learning-detected leads was compared to the averaged Airborne Electromagnetic (AEM)-bird data collected over two days during the CryoSat Validation experiment (CryoVex) field campaign in April 2011. This comparison showed that the proposed machine learning methods had better performance (up to r = 0.83 and Root Mean Square Error (RMSE) = 0.29 m) compared to thickness estimation based on existing lead detection methods (RMSE = 0.86–0.93 m). Sea ice thickness based on the machine learning approaches showed a consistent decline from 2011–2013 and rebounded in 2014.

Landfast sea ice (fast ice) means sea ice that is attached to the shoreline with little or no motion in contrast to pack ice which drifts on the sea. As fast ice plays an important role in the environmental and biological systems of the Antarctic, it is crucial to accurately monitor the spatiotemporal distribution of fast ice. Previous studies on fast ice using satellite remote sensing were mostly focused on the Arctic and near-Arctic areas, whereas few studies were conducted over the Antarctic, especially the West Antarctic region. This research mapped fast ice using multisensor data from 2003 to 2008 based on machine learning approaches - decision trees (DTs) and random forest (RF). A total of seven satellite-derived products, including Advanced Microwave Scanning Radiometer for the Earth observing system brightness temperatures and sea ice concentration, Moderate Resolution Imaging Spectroradiometer (MODIS) ice surface temperature (IST) and Special Sensor Microwave/Imager ice velocity, were used as input variables for identifying fast ice. RF resulted in better performance than that of DT for fast ice classification. Visual comparison of the fast ice classification results with 250-m MODIS images for selected areas also revealed that RF outperformed DT. Ice velocity and IST were identified as the most contributing variables to classify fast ice. Spatiotemporal variations of fast ice in the East and West Antarctic were also examined using the time series of the fast ice maps produced by RF. The residence time of fast ice was much shorter in the West Antarctic than in the East. ⓒ 2015 Taylor & Francisclose0

As convective clouds in Northeast Asia are accompanied by various hazards related with heavy rainfall and thunderstorms, it is very important to detect convective initiation (CI) in the region in order to mitigate damage by such hazards. In this study, a novel approach for CI detection using images from Meteorological Imager (MI), a payload of the Communication, Ocean, and Meteorological Satellite (COMS), was developed by improving the criteria of the interest fields of Rapidly Developing Cumulus Areas (RDCA) derivation algorithm, an official CI detection algorithm for Multi-functional Transport SATellite-2 (MTSAT-2), based on three machine learning approaches—decision trees (DT), random forest (RF), and support vector machines (SVM). CI was defined as clouds within a 16 × 16 km window with the first detection of lightning occurrence at the center. A total of nine interest fields derived from visible, water vapor, and two thermal infrared images of MI obtained 15–75 min before the lightning occurrence were used as input variables for CI detection. RF produced slightly higher performance (probability of detection (POD) of 75.5% and false alarm rate (FAR) of 46.2%) than DT (POD of 70.7% and FAR of 46.6%) for detection of CI caused by migrating frontal cyclones and unstable atmosphere. SVM resulted in relatively poor performance with very high FAR ~83.3%. The averaged lead times of CI detection based on the DT and RF models were 36.8 and 37.7 min, respectively. This implies that CI over Northeast Asia can be forecasted ~30–45 min in advance using COMS MI data.

Melt ponds, a common feature on Arctic sea ice, absorb most of the incoming solar radiation and have a large effect on the melting rate of sea ice, which significantly influences climate change. Therefore, it is very important to monitor melt ponds in order to better understand the sea ice-climate interaction. In this study, melt pond retrieval models were developed using the TerraSAR-X dual-polarization synthetic aperture radar (SAR) data with mid-incidence angle obtained in a summer multiyear sea ice area in the Chukchi Sea, the Western Arctic, based on two rule-based machine learning approaches—decision trees (DT) and random forest (RF)—in order to derive melt pond statistics at high spatial resolution and to identify key polarimetric parameters for melt pond detection. Melt ponds, sea ice and open water were delineated from the airborne SAR images (0.3-m resolution), which were used as a reference dataset. A total of eight polarimetric parameters (HH and VV backscattering coefficients, co-polarization ratio, co-polarization phase difference, co-polarization correlation coefficient, alpha angle, entropy and anisotropy) were derived from the TerraSAR-X dual-polarization data and then used as input variables for the machine learning models. The DT and RF models could not effectively discriminate melt ponds from open water when using only the polarimetric parameters. This is because melt ponds showed similar polarimetric signatures to open water. The average and standard deviation of the polarimetric parameters based on a 15 × 15 pixel window were supplemented to the input variables in order to consider the difference between the spatial texture of melt ponds and open water. Both the DT and RF models using the polarimetric parameters and their texture features produced improved performance for the retrieval of melt ponds, and RF was superior to DT. The HH backscattering coefficient was identified as the variable contributing the most, and its spatial standard deviation was the next most contributing one to the classification of open water, sea ice and melt ponds in the RF model. The average of the co-polarization phase difference and the alpha angle in a mid-incidence angle were also identified as the important variables in the RF model. The melt pond fraction and sea ice concentration retrieved from the RF-derived melt pond map showed root mean square deviations of 2.4% and 4.9%, respectively, compared to those from the reference melt pond maps. This indicates that there is potential to accurately monitor melt ponds on multiyear sea ice in the summer season at a local scale using high-resolution dual-polarization SAR data.

Researchers constantly seek more efficient detection techniques to better utilize enhanced image resolution in accurately detecting and monitoring land cover. Recently, convolutional neural networks (CNNs) have shown high performances comparable to or even better than widely used and adopted machine learning techniques. The aim of this study is to investigate the application of CNNs for land cover classification by using two-dimensional (2-D) spectral curve graphs from multispectral satellite images. The land cover classification was conducted in Concord, New Hampshire, USA, and South Korea by using multispectral images acquired from 30-m Landsat-8 and 500-m Geostationary Ocean Color Imager images. For the construction of input data specific to CNNs, two seasons (winter and summer) of multispectral bands were transformed into 2-D spectral curve graphs for each class. Land cover classification results of CNNs were compared with the results of support vector machines (SVMs) and random forest (RFs). The CNNs model showed higher performance than RFs and SVMs in both study sites. The examination of land cover classification maps demonstrates a good agreement with reference maps, Google Earth images, and existing global scale land cover map, especially for croplands. Using the spectral curve graph could incorporate the phenological cycles on classifying the land cover types. This study shows that the use of a new transformation of spectral bands into a 2-D form for application in CNNs can improve land cover classification performance.