Time Series Of High-resolution Images Research Articles

Fine-scale mapping of Spartina alterniflora-invaded mangrove forests with multi-temporal WorldView-Sentinel-2 data fusion

Mangrove forests, a critical component of blue carbon ecosystems, make a significant contribution to mitigating climate change and maintaining biodiversity. But in China, they have been suffering from severe invasion by Spartina alterniflora (S. alterniflora). Accurate information on spatial distribution of mangroves and S. alterniflora is a prerequisite to protecting indigenous mangrove ecosystems and managing S. alterniflora invasion. However, the classification performance is often constrained by two major issues: (1) A single-date image, regardless of acquisition time, hardly provides optimal spectral separability among evergreen mangroves, deciduous S. alterniflora and mudflats; (2) Due to the gradual invasion process, the mixtures between land cover classes are prevalent in moderate-resolution images. The use of high-resolution (finer than 2 m) image time series could effectively solve these two issues. However, due to frequent cloud cover in coastal areas and long revisit time of high-resolution satellite sensors, the collection of high-resolution image time series is still a great challenge, let alone substantial expenses. In this study, we proposed a new multi-temporal classification approach to fine-scale mapping of exotic S. alterniflora and native mangroves using synthetic high-resolution image time series derived from spatiotemporal fusion of WorldView-2 and Sentinel-2 data. A total of six synthetic WorldView-2-like images in different phenological periods were produced by the Flexible Spatiotemporal DAta Fusion (FSDAF) method. With the combination of the original and synthetic images, both types of image-based multi-temporal features and pixel-based image compositing features were extracted and evaluated in the following Random Forest classifications. Monte Carlo cross-validation was adopted to provide robust accuracy estimates, and a spatially explicit confidence map was generated to show classification uncertainty and potential errors. In order to further demonstrate the necessity of generating synthetic WorldView-2 images, the entire classification process was repeated using Sentinel-2 images at 10-m spatial resolution and the comparison between fine- and moderate-resolution classification results was made in terms of classification accuracies, classification maps and fractional abundances. Our results show that the spectral profiles of S. alterniflora and mangroves generated from the synthetic WorldView-2 images were consistent with the ones from the paired Sentinel-2 images. Meanwhile, the spatial details of high landscape heterogeneity, unrecognizable in the Sentinel-2 images, were revealed by the synthetic WorldView-2 images. Compared to the single-date classification with the original WorldView-2 image, most of multi-temporal classifications using both the original and synthetic images provided more accurate results, with the overall accuracy raised by up to 3.9%. Among various temporal features adopted in this study, the use of the image-based, stacked spectral reflectance and vegetation index features of all original and synthetic images produced the best classification performance, increasing the producer's and user's accuracies of mangroves, S. alterniflora and mudflats by 2.5%–7.1%. Finally, the Sentinel-2 classification accuracies were generally lower than the WorldView-2 counterparts across all temporal features, with a 6% difference in the highest overall accuracy between the two. Prediction errors of the moderate-resolution classification were often found in heterogeneous landscapes, where S. alterniflora was mixed with mangroves or mudflats. Considering the availability of moderate-resolution data with high revisit frequencies, such as Sentinel-2A/B and Landsat 8/9, future fine-scale mapping exercises might benefit from the multi-temporal classification approach with spatiotemporal fusion of moderate- and high-resolution data, alleviating the problem of the scarcity of high-resolution data in cloudy areas.

A 3D CNN APPROACH FOR CHANGE DETECTION IN HR SATELLITE IMAGE TIME SERIES BASED ON A PRETRAINED 2D CNN

Abstract. Over recent decades, Change Detection (CD) has been intensively investigated due to the availability of High Resolution (HR) multi-spectral multi-temporal remote sensing images. Deep Learning (DL) based methods such as Convolutional Neural Network (CNN) have recently received increasing attention in CD problems demonstrating high potential. However, most of the CNN-based CD methods are designed for bi-temporal image analysis. Here, we propose a Three-Dimensional (3D) CNN-based CD approach that can effectively deal with HR image time series and process spatial-spectral-temporal features. The method is unsupervised and thus does not require the complex task of collecting labelled multi-temporal data. Since there are only a few pretrained 3D CNNs available that are not suitable for remote sensing CD analysis, the proposed approach starts with a pretrained 2D CNN architecture trained on remote sensing images for semantic segmentation and develops a 3D CNN architecture using a transfer learning technique to jointly deal with spatial, spectral and temporal information. A layerwise feature reduction strategy is performed to select the most informative features and a pixelwise year-based Change Vector Analysis (CVA) is employed to identify changed pixels. Experimental results on a long time series of Landsat 8 images for an area located in Saudi Arabia confirm the effectiveness of the proposed approach.

Deep Species Distribution Modeling From Sentinel-2 Image Time-Series: A Global Scale Analysis on the Orchid Family.

Species distribution models (SDMs) are widely used numerical tools that rely on correlations between geolocated presences (and possibly absences) and environmental predictors to model the ecological preferences of species. Recently, SDMs exploiting deep learning and remote sensing images have emerged and have demonstrated high predictive performance. In particular, it has been shown that one of the key advantages of these models (called deep-SDMs) is their ability to capture the spatial structure of the landscape, unlike prior models. In this paper, we examine whether the temporal dimension of remote sensing images can also be exploited by deep-SDMs. Indeed, satellites such as Sentinel-2 are now providing data with a high temporal revisit, and it is likely that the resulting time-series of images contain relevant information about the seasonal variations of the environment and vegetation. To confirm this hypothesis, we built a substantial and original dataset (called DeepOrchidSeries) aimed at modeling the distribution of orchids on a global scale based on Sentinel-2 image time series. It includes around 1 million occurrences of orchids worldwide, each being paired with a 12-month-long time series of high-resolution images (640 x 640 m RGB+IR patches centered on the geolocated observations). This ambitious dataset enabled us to train several deep-SDMs based on convolutional neural networks (CNNs) whose input was extended to include the temporal dimension. To quantify the contribution of the temporal dimension, we designed a novel interpretability methodology based on temporal permutation tests, temporal sampling, and temporal averaging. We show that the predictive performance of the model is greatly increased by the seasonality information contained in the temporal series. In particular, occurrence-poor species and diversity-rich regions are the ones that benefit the most from this improvement, revealing the importance of habitat's temporal dynamics to characterize species distribution.

Maize and sorghum field segregation using multi-temporal Sentinel-2 data in central Mexico

Crop segregation in optical remote sensing cannot always be easily achieved among similar species since their spectral response can be similar. We aimed to distinguish fields of two C4 gramineae with similar plant structure and crop calendar, maize (Zea mays) and sorghum (Sorghum bicolor), in an agricultural area of the state of Guanajuato, in central México. One hundred and forty-three fields of these two crops were identified and their average spectral signatures were derived at different phenological stages, using digital information from 11 Sentinel 2A/B images, corresponding to the 2019 spring–summer cycle. We also calculated the temporal profile of six spectral indices for each crop, to evaluate seasonal variations and conducted a cropland classification, using the random forests algorithm and different combinations of dates and indices. Results show that segregation of maize and sorghum fields can be achieved due to small differences in the near-infrared region of their spectral signatures and disparities in the rate of decay during the senescence stage. Classification results also confirmed the separability of these two crops, obtaining a very high overall accuracy and kappa values (98.43% and 97.07%, correspondingly) based on the reflectance of nine images and only three spectral indices: the normalized difference vegetation index, the normalized difference water index, and the structure insensitive pigment index. This investigation demonstrates that the combined use of time series of high-resolution images and spectral indices can contribute to distinguish fields of similar crop species.

A high-resolution image time series of the Gorner Glacier – Swiss Alps – derived from repeated unmanned aerial vehicle surveys

Abstract. Modern drone technology provides an efficient means to monitor the response of alpine glaciers to climate warming. Here we present a new topographic dataset based on images collected during 10 UAV surveys of the Gorner Glacier glacial system (Switzerland) carried out approximately every 2 weeks throughout the summer of 2017. The final products, available at https://doi.org/10.5281/zenodo.2630456 (Benoit et al., 2018), consist of a series of 10 cm resolution orthoimages, digital elevation models of the glacier surface, and maps of ice surface displacement. Used on its own, this dataset allows mapping of the glacier and monitoring surface velocities over the summer at a very high spatial resolution. Coupled with a classification or feature detection algorithm, it enables the extraction of structures such as surface drainage networks, debris, or snow cover. The approach we present can be used in the future to gain insights into ice flow dynamics.

Assessing the impact of different landscape features on post-fire forest recovery with multitemporal remote sensing data: the case of Mount Taygetos (southern Greece)

Fires affecting large areas usually create a mosaic of recovering plant communities reflecting their pre-fire composition and local conditions of burning. However, post-fire recovery patterns may also reveal the effects of landscape heterogeneity on the natural regeneration process of plant communities. This study combines field data and remote sensing image interpretation techniques to assess the role of various landscape characteristics in the post-fire recovery process in a mountainous region of Greece burned by a severe wildfire. Remote sensing techniques were used to accurately map secluded, large burned areas. By introducing a temporal component, we explored the correlation between post-fire regeneration and underlying topography, soils and basement rock. Pre-fire forest cover was reduced by more than half 8 years after fire. Regarding the dominant pre-fire forest trees, Abies cephalonica did not regenerate well after fire and most pre-fire stands were converted to grasslands and shrublands. In contrast, Pinus nigra regenerated sufficiently to return to its pre-fire cover, especially in areas underlain by softer basement rock. The use of different time series of high-resolution images improved the quality of the results obtained, justifying their use despite their high cost.

Time series of high-resolution images enhances efforts to monitor post-fire condition and recovery, Waldo Canyon fire, Colorado, USA

Interpretations of post-fire condition and rates of vegetation recovery can influence management priorities, actions and perception of latent risks from landslides and floods. In this study, we used the Waldo Canyon fire (2012, Colorado Springs, Colorado, USA) as a case study to explore how a time series (2011–2016) of high-resolution images can be used to delineate burn extent and severity, as well as quantify post-fire vegetation recovery. We applied an object-based approach to map burn severity and vegetation recovery using Worldview-2, Worldview-3 and QuickBird-2 imagery. The burned area was classified as 51% high, 20% moderate and 29% low burn-severity. Across the burn extent, the shrub cover class showed a rapid recovery, resprouting vigorously within 1 year, whereas 4 years post-fire, areas previously dominated by conifers were divided approximately equally between being classified as dominated by quaking aspen saplings with herbaceous species in the understorey or minimally recovered. Relative to using a pixel-based Normalised Difference Vegetation Index (NDVI), our object-based approach showed higher rates of revegetation. High-resolution imagery can provide an effective means to monitor post-fire site conditions and complement more prevalent efforts with moderate- and coarse-resolution sensors.

Fusion Approaches for Land Cover Map Production Using High Resolution Image Time Series without Reference Data of the Corresponding Period

Optical sensor time series images allow one to produce land cover maps at a large scale. The supervised classification algorithms have been shown to be the best to produce maps automatically with good accuracy. The main drawback of these methods is the need for reference data, the collection of which can introduce important production delays. Therefore, the maps are often available too late for some applications. Domain adaptation methods seem to be efficient for using past data for land cover map production. According to this idea, the main goal of this study is to propose several simple past data fusion schemes to override the current land cover map production delays. A single classifier approach and three voting rules are considered to produce maps without reference data of the corresponding period. These four approaches reach an overall accuracy of around 80% with a 17-class nomenclature using Formosat-2 image time series. A study of the impact of the number of past periods used is also done. It shows that the overall accuracy increases with the number of periods used. The proposed methods require at least two or three previous years to be used.

Spatially detailed retrievals of spring phenology from single-season high-resolution image time series

Vegetation indices derived from satellite image time series have been extensively used to estimate the timing of phenological events like season onset. Medium spatial resolution (≥250m) satellite sensors with daily revisit capability are typically employed for this purpose. In recent years, phenology is being retrieved at higher resolution (≤30m) in response to increasing availability of high-resolution satellite data. To overcome the reduced acquisition frequency of such data, previous attempts involved fusion between high- and medium-resolution data, or combinations of multi-year acquisitions in a single phenological reconstruction. The objectives of this study are to demonstrate that phenological parameters can now be retrieved from single-season high-resolution time series, and to compare these retrievals against those derived from multi-year high-resolution and single-season medium-resolution satellite data. The study focuses on the island of Schiermonnikoog, the Netherlands, which comprises a highly-dynamic saltmarsh, dune vegetation, and agricultural land. Combining NDVI series derived from atmospherically-corrected images from RapidEye (5m-resolution) and the SPOT5 Take5 experiment (10m-resolution) acquired between March and August 2015, phenological parameters were estimated using a function fitting approach. We then compared results with phenology retrieved from four years of 30m Landsat 8 OLI data, and single-year 100m Proba-V and 250m MODIS temporal composites of the same period. Retrieved phenological parameters from combined RapidEye/SPOT5 displayed spatially consistent results and a large spatial variability, providing complementary information to existing vegetation community maps. Retrievals that combined four years of Landsat observations into a single synthetic year were affected by the inclusion of years with warmer spring temperatures, whereas adjustment of the average phenology to 2015 observations was only feasible for a few pixels due to cloud cover around phenological transition dates. The Proba-V and MODIS phenology retrievals scaled poorly relative to their high-resolution equivalents, indicating that medium-resolution phenology retrievals need to be interpreted with care, particularly in landscapes with fine-scale land cover variability.

Time Series Analysis of Landslide Dynamics Using an Unmanned Aerial Vehicle (UAV)

In this study, we used an Unmanned Aerial Vehicle (UAV) to collect a time series of high-resolution images over four years at seven epochs to assess landslide dynamics. Structure from Motion (SfM) was applied to create Digital Surface Models (DSMs) of the landslide surface with an accuracy of 4–5 cm in the horizontal and 3–4 cm in the vertical direction. The accuracy of the co-registration of subsequent DSMs was checked and corrected based on comparing non-active areas of the landslide, which minimized alignment errors to a mean of 0.07 m. Variables such as landslide area and the leading edge slope were measured and temporal patterns were discovered. Volumetric changes of particular areas of the landslide were measured over the time series. Surface movement of the landslide was tracked and quantified with the COSI-Corr image correlation algorithm but without ground validation. Historical aerial photographs were used to create a baseline DSM, and the total displacement of the landslide was found to be approximately 6630 m3. This study has demonstrated a robust and repeatable algorithm that allows a landslide’s dynamics to be mapped and monitored with a UAV over a relatively long time series.

A Simple Fusion Method for Image Time Series Based on the Estimation of Image Temporal Validity

High-spatial-resolution satellites usually have the constraint of a low temporal frequency, which leads to long periods without information in cloudy areas. Furthermore, low-spatial-resolution satellites have higher revisit cycles. Combining information from high- and low- spatial-resolution satellites is thought a key factor for studies that require dense time series of high-resolution images, e.g., crop monitoring. There are several fusion methods in the bibliography, but they are time-consuming and complicated to implement. Moreover, the local evaluation of the fused images is rarely analyzed. In this paper, we present a simple and fast fusion method based on a weighted average of two input images (H and L), which are weighted by their temporal validity to the image to be fused. The method was applied to two years (2009–2010) of Landsat and MODIS (MODerate Imaging Spectroradiometer) images that were acquired over a cropped area in Brazil. The fusion method was evaluated at global and local scales. The results show that the fused images reproduced reliable crop temporal profiles and correctly delineated the boundaries between two neighboring fields. The greatest advantages of the proposed method are the execution time and ease of use, which allow us to obtain a fused image in less than five minutes.

Observational evidence of Joule heating in some umbral dots

We investigate the effect of the Joule heating (JH) mechanism on the brightness of umbral dots (UDs), proposed by Hamedivafa (2003, A&A, 407, 761). Time series of high-resolution images of UDs in a developed sunspot and in a large pore are used for this purpose. The effect of the JH mechanism is characterized by a specific shape of the temporal variations of UD brightness and area and it is observed better in those UDs where the magnetic field strength is very weak compared to their surroundings. In our observations, about 12% of UDs in the sunspot and about 14% of UDs in the pore show indications of the effective presence of the JH power.