Description: During the 4th International Polar Year 2007–2009 (IPY), it has become increasingly obvious that we need to prepare for a new era in the Arctic. IPY occurred during the time of the largest retreat of Arctic sea ice since satellite observations started in 1979. This minimum in September sea ice coverage was accompanied by other signs of a changing Arctic, including the unexpectedly rapid transpolar drift of the Tara schooner, a general thinning of Arctic sea ice and a double-dip minimum of the Arctic Oscillation at the end of 2009. Thanks to the lucky timing of the IPY, those recent phenomena are well documented as they have been scrutinized by the international research community, taking advantage of the dedicated observing systems that were deployed during IPY. However, understanding changes in the Arctic System likely requires monitoring over decades, not years. Many IPY projects have contributed to the pilot phase of a future, sustained, observing system for the Arctic. We now know that many of the technical challenges can be overcome. The Norwegian projects iAOOS-Norway, POLEWARD and MEOP were significant ocean monitoring/research contributions during the IPY. A large variety of techniques were used in these programs, ranging from oceanographic cruises to animal-borne platforms, autonomous gliders, helicopter surveys, surface drifters and current meter arrays. Our research approach was interdisciplinary from the outset, merging ocean dynamics, hydrography, biology, sea ice studies, as well as forecasting. The datasets are tremendously rich, and they will surely yield numerous findings in the years to come. Here, we present a status report at the end of the official period for IPY. Highlights of the research include: a quantification of the Meridional Overturning Circulation in the Nordic Seas (“the loop”) in thermal space, based on a set of up to 15-year-long series of current measurements; a detailed map of the surface circulation as well as characterization of eddy dispersion based on drifter data; transport monitoring of Atlantic Water using gliders; a view of the water mass exchanges in the Norwegian Atlantic Current from both Eulerian and Lagrangian data; an integrated physical–biological view of the ice-influenced ecosystem in the East Greenland Current, showing for instance nutrient-limited primary production as a consequence of decreasing ice cover for larger regions of the Arctic Ocean. Our sea ice studies show that the albedo of snow on ice is lower when snow cover is thinner and suggest that reductions in sea ice thickness, without changes in sea ice extent, will have a significant impact on the arctic atmosphere. We present up-to-date freshwater transport numbers for the East Greenland Current in the Fram Strait, as well as the first map of the annual cycle of freshwater layer thickness in the East Greenland Current along the east coast of Greenland, from data obtained by CTDs mounted on seals that traveled back and forth across the Nordic Seas. We have taken advantage of the real-time transmission of some of these platforms and demonstrate the use of ice-tethered profilers in validating satellite products of sea ice motion, as well as the use of Seagliders in validating ocean forecasts, and we present a sea ice drift product – significantly improved both in space and time – for use in operational ice-forecasting applications. We consider real-time acquisition of data from the ocean interior to be a vital component of a sustained Arctic Ocean Observing System, and we conclude by presenting an outline for an observing system for the European sector of the Arctic Ocean.

Description: Rapid technological advances in airborne hyperspectral and lidar systems paved the way for using machine learning algorithms to map urban environments. Both hyperspectral and lidar systems can discriminate among many significant urban structures and materials properties, which are not recognizable by applying conventional RGB cameras. In most recent years, the fusion of hyperspectral and lidar sensors has overcome challenges related to the limits of active and passive remote sensing systems, providing promising results in urban land cover classification. This paper presents principles and key features for airborne hyperspectral imaging, lidar, and the fusion of those, as well as applications of these for urban land cover classification. In addition, machine learning and deep learning classification algorithms suitable for classifying individual urban classes such as buildings, vegetation, and roads have been reviewed, focusing on extracted features critical for classification of urban surfaces, transferability, dimensionality, and computational expense.

Description: The integration of imaging spectroscopy and aeromagnetics provides a cost-effective and promising way to extend the initial analysis of a mineral deposit. While imaging spectroscopy retrieves surface spectral information, magnetic responses are used to determine magnetization at both shallower and greater depths using 2D and 3D modeling. Integration of imaging spectroscopy and magnetics improves upon knowledge concerning lithology with magnetic properties, enhances understanding of the geological origin of magnetic anomalies, and is a promising approach for analyzing a prospective area for minerals having a high iron-bearing content. To combine iron diagnostic information from airborne hyperspectral and magnetic data, we (a) used an iron absorption feature ratio to model pseudo-magnetic responses and compare them with the measured magnetic data and (b) estimated the apparent susceptibility along the surface by some equivalent source modeling, and compared them with iron ratios along the surface. For this analysis, a Modified Iron Feature Depth index was developed and compared to the surface geochemistry of the rock samples in order to validate the spectral information of iron. The comparison revealed a linear increase in iron absorption feature depths with iron content. The analysis was performed by empirically modeling the statistical relationship between the diagnostic absorption features of hyperspectral (HS) image spectra of selected rock samples and their corresponding geochemistry. Our results clearly show a link between the spectral absorption features and the magnetic response from iron-bearing ultra/-mafic rocks. The iron absorption feature ratio of Fe3+/Fe2+ integrated with aeromagnetic data (residual magnetic anomaly) allowed us to distinguish main rock types based on physical properties. This separation matches the lithology of the Niaqornarssuit complex, our study area in West Greenland.

Description: Multisensor data fusion has become a hot topic in the remote sensing research community. This is thanks to significant technological advances and the ability to extract information that would have been challenging with a single sensor. However, sensory enhancement requires advanced analysis that enables deep learning. A framework is designed to effectively fuse hyperspectral and lidar data for semantic segmentation in the urban environment. Our work proposes a method of reducing dimensions by exploring the most representative features from hyperspectral and lidar data and using them for supervised semantic segmentation. In addition, we chose to compare segmentation models based on 2D and 3D convolutional operations with two different model architectures, such as U-Net and ResU-Net. All algorithms have been tested with three loss functions: standard Categorical Cross-Entropy, Focal Loss and a combination of Focal Loss and Jaccard Distance—Focal–Jaccard Loss. Experimental results demonstrated that the 3D segmentation of U-Net and ResU-Net with Focal and Focal–Jaccard Loss functions had significantly improved performance compared to the standard Categorical Cross-Entropy models. The results show a high accuracy score and reflect reality by preserving the complex geometry of the objects.

Description: Technological innovations and advanced multidisciplinary research increase the demand for multisensor data fusion in Earth observations. Such fusion has great potential, especially in the remote sensing field. One sensor is often insufficient in analyzing urban environments to obtain comprehensive results. Inspired by the capabilities of hyperspectral and Light Detection and Ranging (LiDAR) data in multisensor data fusion at the feature level, we present a novel approach to the multitemporal analysis of urban land cover in a case study in Høvik, Norway. Our generic workflow is based on bitemporal datasets; however, it is designed to include datasets from other years. Our framework extracts representative endmembers in an unsupervised way, retrieves abundance maps fed into segmentation algorithms, and detects the main urban land cover classes by implementing 2D ResU-Net for segmentation without parameter regularizations and with effective optimization. Such segmentation optimization is based on updating initial features and providing them for a second iteration of segmentation. We compared segmentation optimization models with and without data augmentation, achieving up to 11% better accuracy after segmentation optimization. In addition, a stable spectral library is automatically generated for each land cover class, allowing local database extension. The main product of the multitemporal analysis is a map update, effectively detecting detailed changes in land cover classes.