Multi-temporal Image Analysis Research Articles

LANDSCAPES DYNAMICS OF NORTH-EASTERN UZBEKISTAN IN THE CONTEXT OF CLIMATE AND LAND USE CHANGES

The article examines the causes and results of the dynamics of the Western Tien Shan landscapes in the context of climatic and institutional changes. Based on field studies, analysis of multi temporal satellite images over 30 years and sociological surveys of the local people, changes in the landcover and land use of the region were identified. That includes degradation of glacial and subnival landscapes, an increase in the area and change in the structure of agricultural lands, expansion of recreational infrastructure and growth of recreational load, etc. The results obtained can contribute to the development of recommendations for landscape planning of the region and the resolution of conflicts in environmental management, taking into account the interests and needs of the local people.

Chronicle of a death foretold: The vanishing of an emblematic cultural landscape results in the loss of its unique plant communities

Badlands are peculiar geomorphological formations shaping landscapes of high ecological and cultural value. In the last decades, land reclamation for agricultural purposes and the cessation of traditional land use, such as sheep grazing, led to their decline in extent in many areas. To quantify the changes in badland-related plant communities, we resurveyed badland vegetation in a site of the Crete Senesi (Siena, central Italy) after 16 years (2006–2022), using 48 quasi-permanent vegetation plots and by means of uni- and multivariate analysis of variance. We found an increase in the total vegetation cover of plant communities growing in former bare soil and sparsely vegetated areas, in line with an overall decrease in the extent of bare soil surfaces in the study area, which we highlighted through the analysis of multitemporal satellite images. Pioneer vegetation characterized by the endemic plant Artemisia caerulescens subsp. cretacea changed into ruderal annual grasslands, while former bare soils were colonized by such pioneer vegetation. In contrast, perennial grasslands remained stable. Grasslands with shrubs became more similar to perennial grasslands in species composition. Species richness increased in former bare soils, and using the total vegetation cover as a proxy for successional stages, we found that Shannon diversity and evenness peaked at about 90% of total cover. In all the stages of colonization, short distance dispersal species prevailed, both therophytes (Avena sterilis, Parapholis strigosa) and perennials (Artemisia caerulescens subsp. cretacea, Bromopsis erecta). Long distance dispersal species (Galatella linosyris) started colonizing at about 60% of total vegetation cover, and at high vegetation cover all the functional groups coexisted. Our results confirm that the badland landscapes of southern Tuscany and specialist plant diversity adapted to badlands are vanishing after a diminishing of active land management, suggesting the current ineffectiveness of the Natura 2000 network in their conservation.

Multi-Temporal Image Analysis for Preventive Conservation of Historical Musical Instruments

Artworks need to be constantly monitored to check their state of conservation and to quickly spot the eventual presence of alterations or damages. Preventive conservation is the set of practices employed to reach this goal. Unfortunately, this results generally in a cumbersome process involving multiple analytical techniques. Consequently, methods able to provide a quick preliminary examination of the artworks (e.g., optical monitoring) seem very promising to streamline preventive conservation procedures. We are especially interested in the study of historical wood musical instruments, a kind of artwork particularly subject to mechanical wear since they are both held in museums and also occasionally played in concerts. Our primary goal is to detect possible altered regions on the surface of the instruments early and thus provide the experts some precise indications on where to apply more in-depth examinations to check for potential damages. In this work, we propose an optical monitoring method based on the a-contrario probabilistic framework. Tests were conducted on the “Violins UVIFL imagery” dataset, a collection of UV-induced fluorescence image sequences of artificially altered wood samples and violins. Obtained results showed the robustness of the proposed method and its capability to properly detect altered regions while rejecting noise.

Multitemporal hyperspectral satellite image analysis and classification using fast scale invariant feature transform and deep learning neural network classifier

Multitemporal hyperspectral satellite image analysis and classification using fast scale invariant feature transform and deep learning neural network classifier

A Dual-Branch Deep Learning Architecture for Multisensor and Multitemporal Remote Sensing Semantic Segmentation

Multi-sensor data analysis allows exploiting heterogeneous data regularly acquired by the many available Remote Sensing (RS) systems. Machine- and deep-learning methods use the information of heterogeneous sources to improve the results obtained by using single-source data. However, the State-of-the-Art (SoA) methods analyze either the multi-scale information of multi-sensor multi-resolution images or the time component of image time series. We propose a supervised Deep-Learning (DL) classification method that jointly performs a multi-scale and multi-temporal analysis of RS multi-temporal images acquired by different sensors. The proposed method processes Very-High-Resolution (VHR) images using a Residual Network (ResNet) with a wide receptive field that handles geometrical details and multi-temporal High-Resolution (HR) image using a 3D Convolutional Neural Network (3D-CNN) that analyzes both the spatial and temporal information. The multi-scale and multi-temporal features are processed together in a decoder to retrieve a land-cover map. We tested the proposed method on two multi-sensor and multi-temporal datasets. One is composed of VHR orthophotos and Sentinel-2 multi-temporal images for pasture classification, and another is composed of VHR orthophotos and Sentinel-1 multi-temporal images. Results proved the effectiveness of the proposed classification method.

The Application of Multi Temporal Satellite Data For Land Cover Mapping of Bawean Island, East Java

Land cover dynamics in a small island can be determined using Geographic Information System (GIS) approach based on multitemporal image analysis. This study aimed to classify major land cover types and to map land cover changes of Bawean Island. Two sets of 10 meter resolution satellite data ALOS AVNIR (2010) and Sentinel-2A (2020) were used in this study. Satellite image analysis was carried out through several stages namely image pre-processing including radiometric and geometric correction, supervised image classification and accuracy test. Image classification results from 2010 to 2020 showed a significant change in land cover on Bawean Island. The forest vegetation land cover declined significantly from 13,470.5 Ha in 2010 to 8,543.4 Ha in 2020. Most of the area have been converted into paddy fields and built-up areas. The accuracy test and validation were determined by comparing the 2020 Sentinel image classification results with field observation conducted in 2021. The analysis showed good results with 82.52% overall accuracy and 79.66 Kappa coefficient. Further investigation found that changes in land cover on Bawean Island occured due to the agriculture and infrastucture development.
 Keywords : Geographic Information System (GIS); land cover; satellite images; small island
 
 Copyright (c) 2022 Geosfera Indonesia and Department of Geography Education, University of Jember
 This work is licensed under a Creative Commons Attribution-Share A like 4.0 International License

Multitemporal seagrass carbon assimilation and aboveground carbon stock mapping using Sentinel-2 in Labuan Bajo 2019–2020

Seagrass is considered one of the most effective and efficient natural carbon sinks. In order to fully understand how it can help to adapt and mitigate climate change, information on the variability of the extent of its area, carbon stock, assimilation, sequestration, and potential emission is critical. However, details on spatially and temporally explicit seagrass biomass carbon stock, assimilation and sequestration are lacking in many parts of the world, including Indonesia as the seagrass biodiversity hot spot. To address this issue, this research aims to develop a remote sensing model to map aboveground seagrass carbon stock (AGC) and seagrass carbon assimilation rates (CA) based on multitemporal Sentinel-2 image analysis that can be used to assess the variability of seagrass over times and assess the potential carbon loss and emission reduction by seagrass. Sentinel-2 images with ideal conditions (cloud, haze, sunglint-free) and less ideal ones (cloud and haze-free, but with sunglint) were both used as the basis to develop the model. The research used two regression models, the random forest (RF) regression and stepwise regression (SW), and four input types, i.e., the reflectance bands, deglint bands, Principal Component (PC) bands, and co-occurrence texture bands. Field AGC and CA data estimated from percent cover were used to train and assess the accuracy of each model. Cross-validation was also performed to determine the model consistency when applied to images with different conditions. The results showed that the SW regression model using reflectance and deglint bands produced the most accurate and consistent AGC and CA maps. Applied to a 24-month Sentinel-2 image, from 2019 until 2020, the model showed a consistent pattern where the AGC and CA were lower from September to February and higher in March to August for both 2019 and 2020. Based on the two-year seagrass AGC and CA maps, the monthly average AGCs for 2019 and 2020 were estimated at 28.4 ± 2.7 and 28.4 ± 1.3 g C/m2, respectively, while the monthly average CAs were 6.7 ± 1.4 and 6.6 ± 0.7 Mg C/ha/year. The total AGCs in 2019 and 2020 were 113.8 ± 10.9 and 113.9 ± 5.2 Mg C, and the total CAs were 2648.3 ± 623.4 and 2636.9 ± 281.4 Mg C/year. All these values were from 400.71 ha of seagrass meadows. Finally, the availability of such remote sensing models helps to solve the time and cost-effectiveness challenges in acquiring spatially and temporally extensive seagrass AGC and CA information. Moreover, the research was able to produce the first map of multitemporal seagrass AGC and CA in specific seagrass meadow environments in Indonesia, which can be further improved to map these data rapidly at the national level and will be very useful for future work that implements Tier 3 methodological approach on carbon inventory.

STRUCTURAL AND LITHOLOGICAL MODEL OF THE TURBIV DEPOSIT OF ELUVIAL KAOLINS

The geological data on the Turbiv deposit of eluvial kaolins, which is located in Vinnytsia district of Vinnytsia region near the village of Turbiv and is a part of Glukhivetsko-Turbivsky kaolin district, were generalized. Based on the coordinates and description of the wells, the topography of the base and surface of the eluvial kaolin deposits, as well as the lateral change in their thickness, was studied. It has been established that the surface relief and the thickness of the weathering crust are poorly correlated with each other and are conditioned by the history of the geological and geomorphological development of the territory, in particular, by the development of the Middle Eocene, Miocene and modern hydrographic networks, as a result of which the weathering crust was partially eroded. Data on the vertical zonality of the weathering crust and the material composition of its zones are presented. It has been established that the transitions between the zones are not clear, sometimes there are no separate zones, and the profile of the weathering crust, the lithological composition of its zones are determined by the petrotype of the basement crystalline rocks, their mineral composition and the degree of hypergene transformation. The material composition of eluvial kaolins is characterized. Based on the results of well testing and chemical analysis of kaolins, the lateral distribution of oxides of iron, aluminum, titanium and whiteness was studied. It has been established that the spatial distribution of the whiteness index within the deposit varies widely and correlates with the contents of Fe2O3 and TiO2, or one of the indicated oxides. It has been established that in most of the deposit area, the content of Al2O3 is clearly correlated with the whiteness index, namely, areas with a high content of Al2O3 correspond to areas with high whiteness values and vice versa. The vertical distribution of iron, aluminum, titanium oxides and whiteness kaolin within individual wells was studied. It was found that the whiteness index has a direct correlation with the content of Al2O3 and mainly inverse correlation with Fe2O3 and TiO2, оr with one of the oxides. Information is given on the content of iron oxides and the whiteness of kaolins within different subsoil use areas. Kaolins have been found to be enriched in zircon, ilmenite, monazite and rare earth elements, which can be extracted during field development and kaolin enrichment. The conclusion is made about the complexity of the Turbivsky deposit of eluvial kaolins. Based on the analysis of multi-temporal space images, the dynamics of changes in the area of the disturbed geological environment as a result of the extraction of eluvial kaolins during 1990-2020 was studied.

Adaptation to saline intrusion for agriculture farming transformation in the coastal Ben Tre province, Vietnam

This study aimed to evaluate the adaptive capacity (AC) of agriculture farming transformation to improve for the adaptability by saline intrusion of those culture systems in the coastal Ben Tre province. The primary data was collected by questionnaires in three coastal districts of Ba Tri, Thanh Phu and Binh Dai with 178 households to distribute on agriculture land use transformations. An Object-Based Image Analysis (OBIA) and multi-temporal image analysis approach was developed to detect how LULC changes during 2010-2020 in the coastal Ben Tre province using Landsat TM and OLI data. The MODIS (MOD09 8-day reflectance) data was processed into monthly NDVI maps with the Time Series Product Tool software package and then used to classify regionally common rice crops LULC types using unsupervised classification by ISODATA algorithm. Based on primary data collection, evaluating adaptation measures was assessed the compilation of modification options based on a detailed description and criteria including human, society, infrastructure and natural conditions by standardized a weight ranges of adaptive indicators in the agriculture land use changes. The results detected five agriculture land use transformation including rice-shrimp crop rotation to aquaculture, single rice crop to aquaculture, double rice crop to perennial plant/orchards, double rice crop to rice-shrimp crop rotation, and triple rice crop to double rice crop. The adaptability of these transformations conducted the shrimp-rice crop rotation to aquaculture in Binh Dai of strongly high adaptability. Next is followed by double rice crops to orchards and double rice crops to rice-shrimp crop rotation with high adaptability while the rice-shrimp crop rotation to aquaculture is only moderate in Thanh Phu. The low adaptability levels were identified on single rice crop to aquaculture and triple rice crop to double rice crop transformations due to tolerant condition of saltwater for rice crop cultivation. The studies results can further contribute to support adaptation planning in these sectors by using, developing and streamlining this framework to additional and different socio-ecological contexts.

Boulder Fall Ejecta: Present Day Activity on Mars

AbstractBoulder falls are an archives of recent surface activity on Mars, however, determining how recently they fell remains elusive. Our multitemporal HiRISE image analysis shows that new tracks are characterized by a herringbone‐like ejecta pattern at each boulder bounce that we call boulder fall ejecta (BFE). First systematic survey of BFE revealed ∼4,500 tracks whose total integrated track length is ∼900 km. Our BFE longevity analysis reveals that these tracks likely formed in the last few decades. From few examples we also observed slope streaks originating from BFE tracks, providing additional evidence favoring a dry origin theory. BFE fade in as little as ∼2 to 4 Mars years, whereas, others can persist >6 Mars years suggesting differential weathering. Nearly 30% of BFE are observed in the Cerberus Fossae region, which suggests it is one of the most seismically active regions. Thus, BFE can be used to recognize very recent surface processes on planetary surfaces.

Historical Shoreline Analysis and Field Monitoring at Ennore Coastal Stretch along the Southeast Coast of India

A shoreline change analysis has been carried out for the coastal stretch from Ennore creek to Karungali village located along the southeast coast of India. This 15 km-long coastal stretch had undergone significant changes such as erosion and accretion concerning infrastructure developments and leading to large impact on the livelihood of the community. To assess the shoreline changes, the analysis of multi-temporal satellite images has been carried out. A historical trend is established for the study period from 1991 to 2019. The analysis has been made in three timelines considering various developing activities. There was no significant coastal infrastructure development during 1991 to 1999; however, between 1999 and 2009, a major port, pier, and a groyne field were constructed. Additionally, a port was established between 2009 and 2019. Erosion was observed on the coast from Kattupalli to Karungali at a rate of −16.85 m/yr since 2009, while the coast on the south of Ennore port is accreting at the rate of +12.43 m/yr during the same period. The near-future projection using a linear regression model shows further erosion in the coast under similar conditions. The results of this study provide a baseline data for future anthropogenic activities along this coast.

Analysis and assessment of the use of chernozems of the Kungur forest-steppe based on remote sensing data

Based on the combined analysis of soil maps and multi-temporal satellite images of Landsat–5 and Sentinel–2, an analysis of land use within the areas of chernozem soils of the Kungur forest-steppe was carried out. The boundaries of modern lands are established: arable land, meadows and pastures without subdivision, forests and fallow lands. The assessment of changes in the structure of land use on chernozem soils over the past 30 years is given. The current trend consists in a general reduction in the area of agricultural land used, including the area of arable land on chernozem decreased from 73 to 42% of the area, hayfields and pastures – from 5 to 1%. The land removed from agricultural use occupies an area of 19.8 thousand hectares. On abandoned lands, tree and shrub vegetation develops. 65% of the area of chernozem soils is erosion-hazardous.

Cause and Process Mechanism of Rockslide Triggered Flood Event in Rishiganga and Dhauliganga River Valleys, Chamoli, Uttarakhand, India Using Satellite Remote Sensing and in situ Observations

The catchments of Rishiganga and then Dhauliganga valleys in the Chamoli district of Uttarakhand were impacted by a catastrophic flood triggered due to a massive rockslide, caused by wedge failure on 7th February, 2021. It is estimated that the massive rockslide of ~ 23 million cubic meter volume containing base rock, deposited ice, and snow got detached from the northern slopes of the Trishul mountain range near Ronti Glacier and created a vertical fall of almost 1700 m before severely impacting the Ronti Gad valley located at 1.5 km downstream of Ronti Glacier snout. The huge detached mass of rock and ice (GLIMS ID: G079733E30381N) swiftly moved downstream through the glaciated valley entraining snow, debris, mud on its way, caused rapid fluidization, created massive water/slush waves, and washed away partially or completely the hydel power projects and bridges in its route. It is estimated that ~ 0.93 Peta Joules of potential energy led to the generation of a significant amount of kinetic and thermal energy, good enough to trigger above-mentioned processes. Post-event analysis of high-resolution satellite data shows flood water marks in the valley and on the rock outcrops reaching up to ~ 80–150 m height on the way to Raini Village. The mud and the slush produced through this process led to the formation of a dammed lake and temporarily blocked one of the tributaries of the Rishiganga joining from the northeast. This study provides an insight into the sequence of events as they unfolded, through multi-temporal satellite image analysis, aerial survey, seismological data in conjunction with various other geo-spatial and geo-visualization tools for unraveling the flood event that has happened on February 7, 2021. We also discuss the potential cause of rockslide and the process mechanism of this unique event, causing loss of lives and property besides widespread devastation.

Integrated use of archival aerial photogrammetry, GNSS, and InSAR data for the monitoring of the Patigno landslide (Northern Apennines, Italy)

The morphological changes of unstable areas can be identified using different methodologies that allow repeated surveys over time. The integration between the data obtained from different remote sensing and ground-based techniques, characterized by different coverage, resolution, and precision, allows to describe the kinematic motion of landslides with high accuracy and details. The aim of this work is to monitor the displacements of the Patigno landslide, a deep-seated gravitational slope deformation located in the Northern Apennines (Zeri, Massa Carrara, Italy), using archival aerial photogrammetry (1975–2010), continuous GNSS observations (2004–2018), and multi-temporal InSAR data (2015–2019). The results obtained adopting the different techniques were cross-validated and integrated in order to better explain the kinematics of the landslides: the GNSS data analysis shows horizontal movements of about 43 mm/yr in the S-E direction and vertical deformations of 6.5 mm/yr, in agreement with the average displacement rates obtained from photogrammetry and InSAR processing. The analysis of multi-temporal aerial photogrammetric images allowed us to observe three sectors of the landslide body characterized by different velocities rates and planimetric directions, in agreement with the LOS InSAR displacement field. Furthermore, the correlation between the rainfall distribution and the GNSS time series shows an acceleration of the sliding movements after about 3–4 months of a strong rainfall period. This integrated approach allowed us to overcome the limitations of each technique and to provide a 44-year long monitoring of the Patigno landslide. We also show that a synergic use of ground-based and remote sensing methodologies can provide useful information for the planning of more effective landslide risk mitigation strategies.

Evaluating the temporal accuracy of grassland to cropland change detection using multitemporal image analysis

Grasslands are valuable carbon sinks in the effort to mitigate climate change. However, they are not well protected and are consequently being replaced by agricultural systems worldwide. Current monitoring efforts using remote sensing and ground-based methods are insufficient, and accordingly the mapping of grassland to cropland conversions must be improved to better document these changes in the Canadian Prairies. The purpose of this study is to evaluate different structural break methods and remote sensing datasets for their temporal accuracy in detecting grassland conversions in two Alberta study areas from 2010 to 2018. Breaks For Additive Seasonal and Trend (BFAST), BFAST Seasonal and Bayesian Estimator of Abrupt change, Seasonality and Trend (BEAST) methods were applied to evaluate their sensitivity to rangeland and pasture conversions. The best model was BFAST Seasonal, correctly predicting the year of change for 76% of rangelands and 66% of pastures. This demonstrates that seasonal models are effective in detecting interannual changes in vegetation composition amidst background noise from climate and management induced phenological changes. MODIS data outperformed Landsat, outlining the importance of high temporal resolution remote sensing data to successful change detection, even at the expense of higher spatial resolution. Overall, this study demonstrates that structural break methods are effective in identifying grassland to agriculture transitions and may be useful for the operational monitoring of grassland inventories in the future.

EARTH OBSERVATIONS AS A TOOL FOR DETECTING AND MONITORING POTENTIAL ENVIRONMENTAL VIOLATIONS AND POLICY IMPLEMENTATION

Abstract. In recent years advances in Earth Observation (EO) have increased our ability to inventory, monitor, and understand Earth’s natural and artificial ecosystems. In particular, recent technological developments in the instrument characteristics of the remote sensing systems enabled large-scale usage of EO in environmental governance and protection. Remote sensing can be a potent tool employed in the field of environmental compliance. However, to extract and distribute relevant information to potential stakeholders, simple, yet efficient and robust approaches are needed. The focus of our research is the development of EO algorithms and the corresponding EO processing chains that could be used operationally for providing cost-efficient information to end-users to address environmental compliance and tackle environmental violations over specific thematic domains using freely available Copernicus data. The EO algorithms and the relevant EO processing chains have been developed as part of the EU-H2020 funded project: “EnviroLENS-Copernicus for environmental law enforcement support” where the demonstration area of our approach is in the Montenegro – Albania border. The case considers several sites along the border to Albania, from the coastal area of Velika Plaža to the North Lake Skadar National Park. All sites have been designated in the past as protected areas, due to their rich biodiversity and threatened habitats. The focus of this contribution is on the deforestation detection processing chain. The developed EO processing chains rely on the use and analysis of multi-temporal Sentinel-2 images, while it adopts an object-based perspective for detecting, quantifying, and mapping potential environmental violations.

Area and Mass Changes of Glaciers in the West Kunlun Mountains Based on the Analysis of Multi-Temporal Remote Sensing Images and DEMs from 1970 to 2018

Most of the world’s glaciers have retreated significantly against the background of recent climate warming, while reports have indicated that the glaciers in the West Kunlun Mountains (WKL) may be in a relatively stable state, although there are some gaps in previous research. Based on Landsat series data, topographic maps, SRTM and TanDEM-x data, this paper extracts detailed glacial area information and glacial mass balance during different time periods from 1970 to 2018. We found that, the total area of glaciers in the WKL decreased by 8.0 km2 from 1972 to 2018. The area decreased by 12.0 km2 from 1972 to 1991 and increased by 4.7 km2 from 2010 to 2018. Glacier surface elevation change results in the WKL showed that the overall glacier thickness slightly decreased from 1970 to 2016, with an average of 1.9 ± 1.0 m. The glaciers thinned by approximately 2.5 ± 1.0 m from 1970 to 2000, while from 2000 to 2016, the glaciers thickened approximately by 0.6 ± 1.0 m. Overall, the glaciers in the WKL showed very slight retreat. In addition, the mass changes of glaciers were affected by glacial surging.

AN EXPLAINABLE CONVOLUTIONAL AUTOENCODER MODEL FOR UNSUPERVISED CHANGE DETECTION

Abstract. Transfer learning methods reuse a deep learning model developed for a task on another task. Such methods have been remarkably successful in a wide range of image processing applications. Following the trend, few transfer learning based methods have been proposed for unsupervised multi-temporal image analysis and change detection (CD). Inspite of their success, the transfer learning based CD methods suffer from limited explainability. In this paper, we propose an explainable convolutional autoencoder model for CD. The model is trained in: 1) an unsupervised way using, as the bi-temporal images, patches extracted from the same geographic location; 2) a greedy fashion, one encoder and decoder layer pair at a time. A number of features relevant for CD is chosen from the encoder layer. To build an explainable model, only selected features from the encoder layer is retained and the rest is discarded. Following this, another encoder and decoder layer pair is added to the model in similar fashion until convergence. We further visualize the features to better interpret the learned features. We validated the proposed method on a Landsat-8 dataset obtained in Spain. Using a set of experiments, we demonstrate the explainability and effectiveness of the proposed model.

Multi-temporal image analysis for LULC classification and change detection

ABSTRACT Land use/Land cover (LULC) plays a vital role in planning and supervising the utilization of the natural resources based on the gradual increase in the human demands in the current ecosystem. This study mainly focuses on the usage of the geographic information system (GIS) and land usage to identify the changes in Ananthapuramu, which is located in the Ananthapur district of Andhra Pradesh (AP) state, South India. This paper mainly focuses on the classification and identification of the changes in LULC in the period of 1999–2019 based on a survey of India’s topographic map and temporal images gathered from the satellite. The experimental results indicate that over the specified period of study, the rate of increment and decrement in the built-up area, rate of underwater bodies, forest bodies that may demonstrate a significant impact on the environmental ecosystem.

Wetland Surface Water Detection from Multipath SAR Images Using Gaussian Process-Based Temporal Interpolation

Wetlands provide society with a myriad of ecosystem services, such as water storage, food sources, and flood control. The ecosystem services provided by a wetland are largely dependent on its hydrological dynamics. Constant monitoring of the spatial extent of water surfaces and the duration of flooding of a wetland is necessary to understand the impact of drought on the ecosystem services a wetland provides. Synthetic aperture radar (SAR) has the potential to reveal wetland dynamics. Multitemporal SAR image analysis for wetland monitoring has been extensively studied based on the advances of modern SAR missions. Unfortunately, most previous studies utilized monopath SAR images, which result in limited success. Tracking changes in individual wetlands remains a challenging task because several environmental factors, such as wind-roughened water, degrade image quality. In general, the data acquisition frequency is an important factor in time series analysis. We propose a Gaussian process-based temporal interpolation (GPTI) method that enables the synergistic use of SAR images taken from multiple paths. The proposed model is applied to a series of Sentinel-1 images capturing wetlands in Okanogan County, Washington State. Our experimental analysis demonstrates that the multiple path analysis based on the proposed method can extract seasonal changes more accurately than a single path analysis.