Remote sensing images provide a valuable way to observe the Earth’s surface and identify objects from a satellite or airborne perspective. Researchers can gain a more comprehensive understanding of the Earth’s surface by using a variety of heterogeneous data sources, including multispectral, hyperspectral, radar, and multitemporal imagery. This abundance of different information over a specified area offers an opportunity to significantly improve change detection tasks by merging or fusing these sources. This review explores the application of deep learning for change detection in remote sensing imagery, encompassing both homogeneous and heterogeneous scenes. It delves into publicly available datasets specifically designed for this task, analyzes selected deep learning models employed for change detection, and explores current challenges and trends in the field, concluding with a look towards potential future developments.

A Bayesian classifier mapped the Burned Area (BA) in the Northeastern Siberian boreal forest (70°N 120°E–60°N 170°E) from 1982 to 2015. The algorithm selected the 0.05° (~5 km) Long-Term Data Record (LTDR) version 3 and 4 data sets to generate 10-day BA composites. Landsat-TM scenes of the entire study site in 2002, 2010, and 2011 assessed the spatial accuracy of this LTDR-BA product, in comparison to Moderate-Resolution Imaging Spectroradiometer (MODIS) MCD45A1 and MCD64A1 BA products. The LTDR-BA algorithm proves a reliable source to quantify BA in this part of Siberia, where comprehensive BA remote sensing products since the 1980s are lacking. Once grouped by year and decade, this study explored the trends in fire activity. The LTDR-BA estimates contained a high interannual variability with a maximum of 2.42 million ha in 2002, an average of 0.78 million ha/year, and a standard deviation of 0.61 million ha. Going from 6.36 in the 1980s to 10.21 million ha BA in the 2010s, there was a positive linear BA trend of approximately 1.28 million ha/decade during these last four decades in the Northeastern Siberian boreal forest.

Wildland fires are one of the factors causing the deepest disturbances on the natural environment and severely threatening many ecosystems, as well as economic welfare and public health. Having accurate and up-to-date fuel type maps is essential to properly manage wildland fire risk areas. This research aims to assess the viability of combining Geographic Object-Based Image Analysis (GEOBIA) and the fusion of a WorldView-2 (WV2) image and low density Light Detection and Ranging (LiDAR) data in order to produce fuel type maps within an area of complex orography and vegetation distribution located in the island of Tenerife (Spain). Independent GEOBIAs were applied to four datasets to create four fuel type maps according to the Prometheus classification. The following fusion methods were compared: Image Stack (IS), Principal Component Analysis (PCA) and Minimum Noise Fraction (MNF), as well as the WV2 image alone. Accuracy assessment of the maps was conducted by comparison against the fuel types assessed in the field. Besides global agreement, disagreement measures due to allocation and quantity were estimated, both globally and by fuel type. This made it possible to better understand the nature of disagreements linked to each map. The global agreement of the obtained maps varied from 76.23% to 85.43%. Maps obtained through data fusion reached a significantly higher global agreement than the map derived from the WV2 image alone. By integrating LiDAR information with the GEOBIAs, global agreement improvements by over 10% were attained in all cases. No significant differences in global agreement were found among the three classifications performed on WV2 and LiDAR fusion data (IS, PCA, MNF). These study’s findings show the validity of the combined use of GEOBIA, high-spatial resolution multispectral data and low density LiDAR data in order to generate fuel type maps in the Canary Islands.

To ensure confidence, measurements carried out by imaging radiometers mounted on satellites require robust validation using “fiducial quality” measurements of the same in situ parameter. For surface temperature measurements this is optimally carried out by radiometers measuring radiation emitted in the infrared region of the spectrum, collocated to that of a satellite overpass. For ocean surface temperatures the radiometers are usually on board ships to sample large areas but for land and ice they are typically deployed at defined geographical sites. It is of course critical that the validation measurements and associated instrumentation are internationally consistent and traceable to international standards. The Committee on Earth Observation Satellites (CEOS) facilitates this process and over the last two decades has organized a series of comparisons, initially to develop and share best practice, but now to assess metrological uncertainties and degree of consistency of all the participants. The fourth CEOS comparison of validation instrumentation: blackbodies and infrared radiometers, was held at the National Physical Laboratory (NPL) during June and July 2016, sponsored by the European Space Agency (ESA). The 2016 campaign was completed over a period of three weeks and included not only laboratory-based measurements but also representative measurements carried out in field conditions, over land and water. This paper is one of a series and reports the results obtained when radiometers participating in this comparison were used to measure the radiance temperature of the NPL ammonia heat-pipe blackbody during the 2016 comparison activities (i.e., an assessment of radiometer performance compared to international standards). This comparison showed that the differences between the participating radiometer readings and the corresponding temperature of the reference blackbody were within the uncertainty of the measurements, but there were a few exceptions, particularly for a reference blackbody temperature of −30°C. Reasons that give rise to the discrepancies observed at the low blackbody temperatures were identified.