Moderate Imaging Research Articles

SRSe-Net: Super-Resolution-Based Semantic Segmentation Network for Green Tide Extraction

Due to the phenomenon of mixed pixels in low-resolution remote sensing images, the green tide spectral features with low Enteromorpha coverage are not obvious. Super-resolution technology based on deep learning can supplement more detailed information for subsequent semantic segmentation tasks. In this paper, a novel green tide extraction method for MODIS images based on super-resolution and a deep semantic segmentation network was proposed. Inspired by the idea of transfer learning, a super-resolution model (i.e., WDSR) is first pre-trained with high spatial resolution GF1-WFV images, and then the representations learned in the GF1-WFV image domain are transferred to the MODIS image domain. The improvement of remote sensing image resolution enables us to better distinguish the green tide patches from the surrounding seawater. As a result, a deep semantic segmentation network (SRSe-Net) suitable for large-scale green tide information extraction is proposed. The SRSe-Net introduced the dense connection mechanism on the basis of U-Net and replaces the convolution operations with dense blocks, which effectively obtained the detailed green tide boundary information by strengthening the propagation and reusing features. In addition, the SRSe-Net reducs the pooling layer and adds a bridge module in the final stage of the encoder. The experimental results show that a SRSe-Net can obtain more accurate segmentation results with fewer network parameters.

COMPARISON OF SPLIT WINDOWS ALGORITHM AND PLANCK METHODS FOR SURFACE TEMPERATURE ESTIMATION BASED ON REMOTE SENSING DATA IN SEMARANG

Surface temperature is one of the parameters in land–surface physical processes and is applied to global warming, climate change, and cycle hydrology. Two thermal bands in Landsat 8 imagery can be used as input for surface temperature extraction using the Split Windows Algorithm (SWA) and Planck method. This study aims to compare surface temperature estimates using the SWA and Planck methods and determine the surface temperature distribution based on the condition of land cover and its changes. The remote sensing data used are Landsat-8 OLI/TIRS Aqua MODIS images on August 27, 2013, and October 1, 2020. The results showed that Landsat 8 could obtain land cover information with an accuracy of 90% in 2013 and 91% in 2020. Planck surface temperature in 2013 was 1-3°C higher than SWA, while in 2020, Planck was 0.001-0.05°C lower than SWA but had similar distribution and pattern. The vegetation in the study area's central and south sides has a lower surface temperature than the built-up area on the north side. Land cover changes from non-built up to build-up area cause surface temperatures to increase. Each land cover has a different emissivity value and affects the surface temperature value, i.e., the lower the emissivity, the higher the surface temperature. The emissivity with surface temperature has a pearson correlation value ≥-0.8**.Keywords: Surface Temperature, Split Windows Algorithm, Planck

A deep learning method for cyanobacterial harmful algae blooms prediction in Taihu Lake, China

Cyanobacterial Harmful Algae Blooms (CyanoHABs) in the eutrophic lakes have become a global environmental and ecological problem. In this study, a CNN-LSTM integrated model for predicting the CyanoHABs area was proposed and applied to the prediction of the CyanoHABs area in Taihu Lake. Firstly, the time-series data of the CyanoHABs area in Taihu Lake for 20 years were accurately obtained using MODIS images from 2000 to 2019 based on the FAI method. Then, a principal component analysis was performed on the daily meteorological data for the month before the outbreak of CyanoHABs in Taihu Lake from 2000 to 2019 to determine the meteorological factors closely related to the outbreak of CyanoHABs. Finally, the features of CyanoHABs area and meteorological data were extracted by Convolutional Neural Networks (CNN) model and used as the input of Long Short Term Memory Network (LSTM). An integrated CNN-LSTM model approach was constructed for predicting the CyanoHABs area. The results show that high R2 (0.91) and low mean relative error (17.42%) verified the validity of the FAI index to extract the CyanoHABs area in Taihu Lake; the meteorological factors closely related to the CyanoHABs outbreak in Taihu Lake are mainly temperature, relative humidity, wind speed, and precipitation; the CNN-LSTM integrated model has better prediction effect for both training and test sets compared with the CNN and LSTM models. This study provides an effective method for predicting temporal changes in the CyanoHABs area and offers new ideas for scientific and effective regulation of inland water safety.

Long-Term Series of Chlorophyll-a Concentration in Brazilian Semiarid Lakes from Modis Imagery

By monitoring the chlorophyll a concentration (chla), it is possible to keep track of the eutrophication status of a lake and to describe the temporal dynamics of the phytoplankton biomass. Such monitoring must be both extensive and intensive to account for the short- and long-term biomass variations. This may be achieved by the remote estimation of chla through an orbital sensor with high temporal resolution. In this study, we used MODIS imagery to produce 21-year time series of chla for three strategic lakes of the Brazilian semi-arid region: Eng. Armando Ribeiro Gonçalves, Castanhão, and Orós. We used data collected in 13 lakes of the region to test new and published regression models for chla estimation. The selected model was validated and applied to daily MODIS images for the three largest lakes. The resulting chla time series revealed that the temporal dynamics of the phytoplankton biomass is associated with the hydraulic regime of the lakes, with chla plummeting upon intense water renewal and keeping high during persistent dry periods. The intense rainy season of 2004 reduced the phytoplankton biomass and its effects even extended to the subsequent years. Our results encourage the exploration of the MODIS archived imagery in limnological studies.

Tropical cyclone effects on vegetation resilience in the Yucatan Peninsula, México, between 2000-2012

The resilience capacity of vegetation in the Yucatan Peninsula is influenced by the winds and rains of tropical cyclones. There are no recent long-term studies on cyclonic impacts on natural vegetation in the region despite their significant effects on infrastructure and biodiversity. The objective of this study was to identify the area impacted by 21 tropical cyclones between 2000 and 2012 and to quantify the recovery capacity of the vegetation by using standardized anomalies of the normalized vegetation index (aNDVI). MODIS images from NASA’s “Terra and Aqua” satellites were used to calculate the damaged areas by analyzing the frequency of pixels corresponding to each type of vegetation per impact zone. The results showed that in 67% of the tropical cyclones, the impacts on vegetation were negative —a decrease in aNDVI—but in 33% of the cyclones, positive effects were recorded —an increase in aNDVI—. The lapse rate of vegetation recovery varied in 52% of the cases; vegetation recovered between two and three weeks after each cyclonic event, while 38% of the cases recovered within four to five weeks of the cyclone landfall. Tropical forests suffered the most significant effects, followed by hydrophilic vegetation. The most destructive hurricanes were Emily, Wilma, and Dean. The rate of recovery laps ranged from 4 to 10 weeks after the hurricane hit. The results could improve assessments of vegetation vulnerability against severe hydrometeorological events and establish priority zones for prompt inspection.

Climate-driven variations in suspended particulate matter dominate water clarity in shallow lakes.

Secchi disk depth (SDD) has long been considered as a reliable proxy for lake clarity, and an important indicator of the aquatic ecosystems. Meteorological and anthropogenic factors can affect SDD, but the mechanism of these effects and the potential control of climate change are poorly understood. Preliminary research at Lake Khanka (international shallow lake on the China-Russia border) had led to the hypothesis that climatic factors, through their impact on suspended particulate matter (SPM) concentration, are key drivers of SDD variability. To verify the hypothesis, Landsat and MODIS images were used to examine temporal trend in these parameters. For that analysis, the novel SPM index (SPMI) was developed, through incorporation of SPM concentration effect on spectral radiance, and was satisfactorily applied to both Landsat (R2 = 0.70, p < 0.001) and MODIS (R2 = 0.78, p < 0.001) images to obtain remote estimates of SPM concentration. Further, the SPMI algorithm was successfully applied to the shallow lakes Hulun, Chao and Hongze, demonstrating its portability. Through analysis of the temporal trend (1984-2019) in SDD and SPM, this study demonstrated that variation in SPM concentration was the dominant driver (explaining 63% of the variation as opposed to 2% due to solar radiation) of SDD in Lake Khanka, thus supporting the study hypothesis. Furthermore, we speculated that variation in wind speed, probably impacted by difference in temperature between lake surface and surrounding landscapes (greater difference between 1984-2009 than after 2010), may have caused varying degree of sediment resuspension, ultimately controlling SPM and SDD variation in Lake Khanka.

Daily monitoring algorithms to detect geostationary imager visible radiance anomalies

The NASA Clouds and the Earth’s Radiant Energy System (CERES) project provides observed flux and cloud products for the climate science community. Geostationary satellite (GEO) imager measured clouds and broadband derived fluxes are incorporated in the CERES SYN1deg product to provide regional diurnal information in between Sun-synchronous Terra and Aqua CERES measurements. The recently launched GEO imagers with onboard calibration systems have active calibration teams that incrementally update the calibration in order to mitigate calibration drifts. However, short-term L1B radiance anomalies and calibration adjustment discontinuities may still exist in the record. To avoid any GEO cloud and flux artifacts in the CERES SYN1deg product, these calibration events must be addressed while scaling the GEO imagers to the Aqua-moderate resolution imaging spectroradiometer (MODIS) calibration reference. All-sky tropical ocean ray-matching (ATO-RM) and deep convective cloud invariant target (DCC-IT)-based monitoring algorithms are presented to detect calibration-driven daily anomalies in the GOES-16 Advanced Baseline Imager L1B visible (0.65 μm) radiance measurements. Sufficient daily ATO-RM sampling was obtained both by ray-matching GOES-16 with multiple MODIS and visible-infrared imaging radiometer suite imagers as well as by increasing the grid resolution. Optimized angular matching and outlier filtering were most effective in reducing the ATO-RM daily gain algorithm noise. The DCC-IT daily calibration algorithm utilized a larger domain and included more GOES-16 scan times. The DCC-IT daily gain uncertainty was reduced by normalizing the DCC regional reflectance on a regional, seasonal, and diurnal basis. The combination of ATO-RM and DCC-IT daily monitoring algorithms is shown to detect, with a high degree of confidence, daily GOES-16 L1B calibration-driven radiance anomalies >2.4 % , while keeping false positives at a minimum. Remarkably, the ATO-RM and DCC-IT daily gains are mostly within 0.5%. The ATO-RM and DCC-IT daily monitoring algorithms can be easily adapted to other GEO imagers and visible channels.

Monitoring irrigation dynamics in paddy fields using spatiotemporal fusion of Sentinel-2 and MODIS

Rapid and accurate monitoring of irrigation dynamics in paddy fields (the start, end, duration and irrigation peak, etc.) at field scales is crucial to the fine management of agricultural water resources, especially in typical areas with water shortages. However, there is still a lack of sufficient research to depict irrigation dynamics in paddy fields at high temporal and spatial levels. To this end, this study fused Sentinel-2 and MODIS images to map the spatio-temporal dynamics of irrigation events in paddy fields. A popular spatiotemporal fusion algorithm (enhanced spatial and temporal adaptive reflectance fusion model, ESTARFM) was used to generate 25 high-spatial resolution (10 m) remote sensing images based on 9 Sentinel-2 images and 24 MODIS images. Random forest algorithm was used to extract the spatial distribution of irrigated paddy fields. Water body index and vegetation index were employed to identify the start, end and duration of irrigation in paddy fields. Penman-Monteith model was used to estimate water surplus and deficit of irrigation during the critical irrigation period with daily observation data from meteorological stations. This study was carried out in rice-growing areas in the middle and lower reaches of the Yellow River in China. The results indicated that the spatial distribution difference of irrigation events in paddy fields with the shortest 3-day interval could be monitored in collaboration with Sentinel-2 and MODIS images. The start, end, and duration of irrigation in paddy fields presented significant spatial differences at field scales. A large amount of water from groundwater and Yellow River was needed, because the total water shortage of irrigation in paddy fields in the study area was about 80.79% of the total water demand. In addition, paddy fields with higher water demand were more concentrated spatially, as were paddy fields with lower water demand. The feasibility of spatiotemporal fusion of multi-source remote sensing data makes it possible to continuously monitor irrigation dynamics in paddy fields on high spatial resolution scales, which is conducive to the construction of spatiotemporal database and big data platform of agricultural irrigation information. This would not only help in promoting the high-quality development of agricultural water resources management but also alleviating the contradiction of regional water resources.

Subpixel Change Detection Based on Improved Abundance Values for Remote Sensing Images

To achieve land cover change detection (LCCD) with both fine spatial and temporal resolutions from remote sensing images, subpixel mapping-based approaches have been widely studied in recent years. The fine spatial but coarse temporal resolution image and the coarse spatial but fine temporal image are used to accomplish LCCD by combining their advantages. However, the performance of subpixel mapping is easily affected by the accuracy of spectral unmixing, thereby reducing the reliability of LCCD. In this paper, a novel subpixel change detection scheme based on improved abundance values is proposed to tackle the aforementioned problem, in which the spatial distribution of fine spatial resolution image is borrowed to promote the accuracy of spectral unmixing. First, the coarse spatial resolution image is used to generate the original abundance image by the spectral unmixing method. Second, the spatial distribution information of the fine spatial resolution image is incorporated into the original abundance image to obtain improved abundance values. Third, the fine spatial resolution subpixel map can be generated by the subpixel mapping method using the improved abundance values. At last, the fine resolution change map can be obtained by comparing the subpixel map with the fine spatial resolution image. Experiments are conducted on a simulated dataset based on Landsat-7 images and two real datasets based on Landsat-8 and MODIS images. The results of the real datasets showed that the proposed method can effectively improve the performance of LCCD with an overall accuracy of approximately 1.26% and 0.79% to the existing methods.

Generalized Linear Spectral Mixing Model for Spatial–Temporal–Spectral Fusion

Image fusion effectively solves the trade-off between spatial resolution, temporal resolution, and spectral resolution of remote sensing sensors. However, most of existing methods focus on the fusion of two of the spatial, temporal, and spectral metrics of remote sensing images. The few spatial-temporal-spectral fusion (STSF) methods available are mainly for fusing MODIS and Landsat images, which are not suitable for the characteristics of the spaceborne hyperspectral images with low temporal resolution, such as Hyperion, ZY-1 02D, and PRISMA. For this purpose, we proposed a novel generalized linear spectral mixing model for spatial-temporal-spectral fusion (GLMM-STSF). In the method, the GLMM is introduced into the STSF problem, and the temporal variations of images at different times are transferred to the endmember and abundance matrix variations of images for estimation. To the best of our knowledge, for the first time, the STSF task of remote sensing images is handled from the perspective of spectral unmixing. Compared with existing STSF fusion methods, our method targets the task of fusing spaceborne HSI with low temporal and spatial resolutions with multispectral image featured by high temporal and spatial resolutions. Taking the STSF of ZY-1 02D hyperspectral and Sentinel-2 multispectral real datasets as an example, comparisons with related state-of-the-art methods demonstrate that our proposed method achieves superior fusion performance.

A Robust Model for MODIS and Landsat Image Fusion Considering Input Noise

Significant progress has been made in spatiotemporal fusion for remote sensing images; however, most models require inputs to be free of clouds and without missing data, considerably confining their applications in practice. Due to recent advances in deep learning technologies, powerful modeling capabilities could be leveraged to bring potential solutions to this problem. This article proposes a novel architecture named the robust spatiotemporal fusion network (RSFN) based on the generative adversarial network and attention mechanism with dual temporal references to automatically handle input noise. The RSFN only needs one coarse-resolution image on the prediction date and two referential fine-resolution images before and after the prediction date as model inputs. Most notably, there is no special restriction attached on the data quality of referential images. The comparison with other models demonstrates the effectiveness of the RSFN model quantitatively and visually in four study areas using MODIS and Landsat images. Two main conclusions can draw from the experiments. First, the input data noise hardly affects the prediction results of the RSFN, and the RSFN can gain a comparable or even higher accuracy; conversely, the other methods only show limited resistance to input noise. Second, the RSFN with cloud-contaminated references outperforms the other models with cloud-free references after data filtering in the same study area during the same period. The satellite data quality usually varies significantly; the model robustness and fault tolerance are considered critical for actual applications. The RSFN is a simple end-to-end deep model with high accuracy and fault tolerance designed for spatiotemporal fusion with imperfect data inputs, showing promising prospects in practical applications.

Machine Learning Based Analysis of Real-Time Geographical of RS Spatio-Temporal Data

Flood disasters can be reliably monitored using remote sensing photos with great spatiotemporal resolution. However, satellite revisit periods and extreme weather limit the use of high spatial resolution images. As a result, this research provides a method for combining Landsat and MODIS pictures to produce high spatiotemporal imagery for flood disaster monitoring. Using the spatial and temporal adaptive reflectance fusion model (STARFM), the spatial and temporal reflectance unmixing model (STRUM), and three prominent algorithms of flexible spatiotemporal data fusion (FSDAF), Landsat fusion images are created by fusing MODIS and Landsat images. Then, to extract flood information, utilize a support vector machine (SVM) to classify the fusion images. Assess the accuracy of your work. Experimental results suggest that the three spatio-temporal fusion algorithms may be used to effectively monitor floods, with FSDAF's fusion results outperforming STARFM and STRUM in both study areas. The overall flood classification accuracy of the three STARFM, STRUM, and FSDAF algorithms in the Gwydir research region is 0.89, 0.90, and 0.91, respectively, with Kappa coefficients of 0.63, 0.64, and 0.67. The flood classification accuracy of the three fusion algorithms in the New Orleans research region is 0.90, 0.89, and 0.91, with Kappa values of 0.77, 0.76, and 0.81, respectively. The spatio-temporal fusion technique can be used to successfully monitor floods, according to this study.

Pre and Post Effects Assessment of Marine Ranch Construction in Chlorophyll-a Concentration Using MODIS Data and a Web-Based Tool. A Case Study in Zhelin Bay, China

Chlorophyll-a (Chl-a) concentration in lakes can tell a lot about a lake’s water quality and ecosystem. It is a measure of the amount of algae growing in a waterbody and can be used to monitor the trophic condition of a waterbody. We studied the pre and post effects of marine ranch construction in Chl-a concentration in Zhelin Bay, Southern China using Normalized Difference Chlorophyll Index (NDCI) and a web-based tool (https://mapcoordinates.info/). We used 8 day composite MODIS image collections of 500 m resolution and randomly selected two stations to extract the chlorophyll-a concentration values through the web-based tool. We recorded the slight increase in NDCI values in all stations after the construction of marine ranch which is a good indicator of the marine organisms’ reproduction and survival.

MODIS Green-Tide Detection With a Squeeze and Excitation Oriented Generative Adversarial Network

This paper presents a novel framework combining spectral analysis and machine learning for green-tide detection. The framework incorporates a squeeze and excitation (SE) attention module into a U-shaped generator of a generative adversarial network (GAN), and is referred to as squeeze and excitation oriented generative adversarial network (SE-GAN). In the SE-GAN, the normalized differential vegetation index (NDVI) is used as the preprocessing filter, which enhances the information associated with green-tide. The SE attention module recalibrates the feature maps so as to enhance the useful features conveyed from the generator&#x2019;s convolution layer while suppressing less useful ones. Overall, the generator attempts to render images that contain green-tide in a way that highly approximates the reference images, while the discriminator characterizes the difference between the generated images and the reference images. The training scheme, which is adversarial, minimizes the <i>f-</i> divergence between the generator and discriminator. Consequently, compared to other green-tide detection algorithms only applicable in small-area scenes, SE-GAN can automatically detect green-tide in MODIS images of any size. Experiments with both large- and small-format MODIS imagery confirm that SE-GAN&#x2019;s detection results are superior to those of five other commonly used methods.

Ekstraksi Perubahan Tutupan Vegetasi Di Kabupaten Gorontalo Menggunakan Google Earth Engine

Monitoring changes in vegetation cover is important for the restoration of ecosystems in the Gorontalo Regency area. The utilization of remote sensing technology makes it possible to detect the dynamics of changes in vegetation cover spatially and temporally. The Terra MODIS satellite image collection in the study area is available in large numbers and sizes. Therefore, cloud computing-based spatial technology support is needed. Google Earth Engine (GEE) as a geospatial computing device is an alternative to cover this shortfall. The aim of this study is to explore the condition of vegetation cover spatially and temporally using the GEE platform. A total of 43 MODIS images in the study area, recording periods 2000 and 2020, were used to quickly and effectively generate vegetation cover maps. The process of downloading, processing, and analyzing data was automated through the GEE interface. The results of the mapping in 2000 and 2020 are shown by maps of vegetation cover in two classes, namely; vegetation and non-vegetation. The accuracy of the vegetation cover map shows good results, namely an overall accuracy of 0.81 for 2000 and 0.85 for 2020. The area of the non-vegetation class increased by 2815.29 ha, and the vegetation class decreased by 2767.31 ha. The map of spatial changes in vegetation cover in the study area is classified into three classes, namely revegetation, devegetation, and unchanged. Based on these results, the extraction of vegetation cover changes in the study area using the GEE platform can be carried out well.

A wildfire growth prediction and evaluation approach using Landsat and MODIS data

The increasing wildfire occurrence due to global climate changes urged the improvement of present wildfire growth prediction and evaluation methods. This study aimed to propose novel solutions to their two primary limitations, including the lack of robust fuel classification method and the low spatial resolution of wildfire growth accuracy assessment while ensuring wide applicability using open data satellite missions and software. The first objective was to create a robust two-step fuel model classification method consisted of the supervised machine learning classification of generalized land cover classes in the 1st level and their individual unsupervised classification to vegetation subtypes in the 2nd level. The second objective was creating a wildfire prediction accuracy assessment method using MODIS 250 m images, which overcome the limitations of low spatial resolution while preserving sub-daily temporal resolution. The wildfire on the Korčula island in Croatia was analyzed in the study, being specific for its long duration from 18 to 24 July 2015. The wildfire ignition occurred in the isolated area, which prolonged the response time from emergency agencies. Random Forest (RF) with input Landsat 8 spectral bands and indices resulted in the highest classification accuracy in the 1st classification level with an overall agreement of 83.6%. The vegetation subclasses from the 2nd classification level were matched to the 13 standard fuel models for the input in FARSITE software. The predicted wildfire evaluation showed the highest mean accuracy of 0.906 for the first two days, which decreased to 0.722 in the latter stages of the active wildfire caused by overprediction. The proposed two-step fuel model classification presented a cost-efficient solution to the fuel map creation in any part of the world, with a disadvantage of no in-situ ground truth identification and accuracy assessment for 2nd classification level. The evaluation of wildfire growth prediction with 250 m images enabled high spatial and temporal resolution of the assessment, while its limitations of wildfire overprediction and the negative effects of wildfire smoke in MODIS images should be addressed in future research.

A Long-term Spatial and Temporal Analysis of NDVI Changes in Java Island Using Google Earth Engine

Java is Indonesia’s and the world’s most populous island. The increase in population on the island of Java reduces the area of forest and other vegetation covers. Landslides, floods, and other natural disasters are caused by reduced vegetation cover. Furthermore, it has the potential to lead to the extinction of flora and fauna. The Normalized Difference Vegetation Index (NDVI) can be used to monitor the vegetation cover. This study analyzes the NDVI changes value from 2005 to 2020 using Terra and Aqua MODIS image data processed using Google Earth Engine. Processing was carried out in some stages: down-setting, performing NDVI processing, calculating monthly average NDVI, calculating annual average NDVI, and analyzing. From the study results, the NDVI value of Terra and Aqua MODIS data has a solid but imperfect correlation coefficient due to differences in orbital time which causes differences in solar zenith angle, sensor viewing angle, and azimuth angle. Then from this study, it was found that overall, changes in vegetation density cover on the island of Java decreased, which was indicated by the NDVI decline rate of -0.00047/year. The most significant decrease in NDVI value occurred in the period 2015–2016, covering an area of 13994.630 km2, and the most significant increase in NDVI occurred in the period 2010–2011, covering an area of 2256.101 km2.

An Investigation on Shenzhen Urban Green Space Changes and Their Effect on Local Eco-Environment in Recent Decades

Rapid urbanization and population growth impact enormous pressures on urban natural, economic and social environments. The quantitative analysis of urban green space (UGS) landscape dynamics and their impact on the urban eco-environment is of great significance for urban planning and eco-environment protection. Taking Shenzhen as an example, the UGS landscape changes and their impact on urban heat islands (UHI), surface wetness, air pollution and carbon storage were comprehensively investigated with Landsat and MODIS images. Results showed a large number of lands transferring from UGS to non-UGS from 1978 to 2018, especially for cropland. Built-up regions have adverse influences on eco-environment factors, and then they suffer high SUHI and AOD and low humidity and carbon storage. The growth of built-up areas not only enlarges the area of SUHI, but also enhances the intensity of heat islands. On the contrary, UGS patches have beneficial influences on all eco-environment factors and then enjoy a better eco-environment, including low SUHII, high surface wetness, high carbon storage and low AOD. It is expected that this study could provide scientific support for UGS plans and for conserving and sustainable urban development for developing cities.

Spatial Downscaling of MODIS Snow Cover Observations Using Sentinel-2 Snow Products

Understanding those processes in which snow dynamics has a significant influence requires long-term and high spatio-temporal resolution observations. While new optical space-borne sensors overcome many previous snow cover monitoring limitations, their short temporal length limits their application in climatological studies. This work describes and evaluates a probabilistic spatial downscaling of MODIS snow cover observations in mountain areas. The approach takes advantage of the already available high spatial resolution Sentinel-2 snow observations to obtain a snow probability occurrence, which is then used to determine the snow-covered areas inside partially snow-covered MODIS pixels. The methodology is supported by one main hypothesis: the snow distribution is strongly controlled by the topographic characteristics and this control has a high interannual persistence. Two approaches are proposed to increase the 500 m resolution MODIS snow cover observations to the 20 m grid resolution of Sentinel-2. The first of these computes the probability inside partially snow-covered MODIS pixels by determining the snow occurrence frequency for the 20 m Sentinel-2 pixels when clear-sky conditions occurred for both platforms. The second approach determines the snow probability occurrence for each Sentinel-2 pixel by computing the number of days in which snow was observed on each grid cell and then dividing it by the total number of clear-sky days per grid cell. The methodology was evaluated in three mountain areas in the Iberian Peninsula from 2015 to 2021. The 20 m resolution snow cover maps derived from the two probabilistic methods provide better results than those obtained with MODIS images downscaled to 20 m with a nearest-neighbor method in the three test sites, but the first provides superior performance. The evaluation showed that mean kappa values were at least 10% better for the two probabilistic methods, improving the scores in one of these sites by 25%. In addition, as the Sentinel-2 dataset becomes longer in time, the probabilistic approaches will become more robust, especially in areas where frequent cloud cover resulted in lower accuracy estimates.

Spatial distribution and interannual variability of cyanobacteria blooms on the North-Western shelf of the Black Sea in 1985-2019 from satellite data

Long-term high and medium resolution satellite data are used to study spatial, seasonal and interannual variability of the cyanobacteria blooms (CB) in the North-Western Shelf (NWS) of the Black Sea. High-resolution Landsat data demonstrate that CB were regularly observed in the study area on satellite images in 1985-2020. Comparison of quasi-synchronous Landsat and MODIS images was used to study the characteristic features of CB spectra, which are: a “step” formed by MODIS 3 (469) and 4 (488) channel, probably related to the absorption by photoprotective carotenoids; increased reflectance in near-infrared; maximum at 547-555 nm. On the base of these features, we develop the automated algorithm of CB identification and reconstruct the daily dataset of CB over 2003-2019. The main advantage of the developed algorithm is its ability to filter out non-bloom areas in all seasons and meteorological conditions and during blooms of other algae in the deep part of the sea. Analysis of this dataset, complemented by Landsat data, allows to identify two primary sources of the CB in the NWS: the Dnieper-Bug estuary and the Danube mouth, and study the evolution of CB in these areas in detail. CB can penetrate on more than 150 km from the sources reaching in rare cases the Crimean coast. Such offshore propagation is mainly wind-driven and was often related to the upwelling frontal currents. The climatic seasonal peak of CB near the Danube is observed in May-June following the peak of river discharge in April-May. It precedes the peak of CB near the Dnieper-Bug estuary observed during maximum surface heating in August. At the same time, the case study of bloom evolution in 2009 shows that CB can be observed in these areas several times in a season, triggered mainly by periods of low winds. The strongest bloom of CB in 2003-2019 was observed in 2004-2006 and 2008-2010 following by the sharp decrease after 2011. This decrease is related to the observed reduction of the river discharge and, more important, the rise of the wind intensity in summers of the recent period.