Spatial-temporal Consistency Research Articles

Mapping global land cover in 2001 and 2010 with spatial-temporal consistency at 250 m resolution

Global land cover types in 2001 and 2010 were mapped at 250m resolution with multiple year time series Moderate Resolution Imaging Spectrometer (MODIS) data. The map for each single year was produced not only from data of that particular year but also from data acquired in the preceding and subsequent years as temporal context. Slope data and geographical coordinates of pixels were also used. The classification system was derived from the finer resolution observation and monitoring of global land cover (FROM-GLC) project. Samples were based on the 2010 FROM-GLC project and samples for other years were obtained by excluding those changed from 2010. A random forest classifier was used to obtain original class labels and to estimate class probabilities for 2000–2002, and 2009–2011. The overall accuracies estimated from cross validation of samples are 74.93% for 2001 and 75.17% for 2010. The classification results were further improved through post processing. A spatial-temporal consistency model, Maximum a Posteriori Markov Random Fields (MAP-MRF), was first applied to improve land cover classification for each 3 consecutive years. The MRF outputs for 2001 and 2010 were then processed with a rule-based label adjustment method with MOD44B, slope and composited EVI series as auxiliary data. The label adjustment process relabeled the over-classified forests, water bodies and barren lands to alternative classes with maximum probabilities.

How can we use MODIS land surface temperature to validate long-term urban model simulations?

High spatial resolution urban climate modeling is essential for understanding urban climatology and predicting the human health impacts under climate change. Satellite thermal remote-sensing data are potential observational sources for urban climate model validation with comparable spatial scales, temporal consistency, broad coverage, and long-term archives. However, sensor view angle, cloud distribution, and cloud-contaminated pixels can confound comparisons between satellite land surface temperature (LST) and modeled surface radiometric temperature. The impacts of sensor view angles on urban LST values are investigated and addressed. Three methods to minimize the confounding factors of clouds are proposed and evaluated using 10 years of Moderate Resolution Imaging Spectroradiometer (MODIS) data and simulations from the High-Resolution Land Data Assimilation System (HRLDAS) over Greater Houston, Texas, U.S. For the satellite cloud mask (SCM) method, prior to comparison, the cloud mask for each MODIS scene is applied to its concurrent HRLDAS simulation. For the max/min temperature (MMT) method, the 50 warmest days and coolest nights for each data set are selected and compared to avoid cloud impacts. For the high clear-sky fraction (HCF) method, only those MODIS scenes that have a high percentage of clear-sky pixels are compared. The SCM method is recommended for validation of long-term simulations because it provides the largest sample size as well as temporal consistency with the simulations. The MMT method is best for comparison at the extremes. And the HCF method gives the best absolute temperature comparison due to the spatial and temporal consistency between simulations and observations.

Comparison of Gross Primary Productivity Derived from GIMMS NDVI3g, GIMMS, and MODIS in Southeast Asia

Gross primary production (GPP) plays an important role in the net ecosystem exchange of CO2 between the atmosphere and terrestrial ecosystems. It is particularly important to monitor GPP in Southeast Asia because of increasing rates of tropical forest degradation and deforestation in the region in recent decades. The newly available, improved, third generation Normalized Difference Vegetation Index (NDVI3g) from the Global Inventory Modelling and Mapping Studies (GIMMS) group provides a long temporal dataset, from July 1981 to December 2011, for terrestrial carbon cycle and climate response research. However, GIMMS NDVI3g-based GPP estimates are not yet available. We applied the GLOPEM-CEVSA model, which integrates an ecosystem process model and a production efficiency model, to estimate GPP in Southeast Asia based on three independent results of the fraction of photosynthetically active radiation absorbed by vegetation (FPAR) from GIMMS NDVI3g (GPPNDVI3g), GIMMS NDVI1g (GPPNDVI1g), and the Moderate Resolution Imaging Spectroradiometer (MODIS) MOD15A2 FPAR product (GPPMOD15). The GPP results were validated using ground data from eddy flux towers located in different forest biomes, and comparisons were made among the three GPPs as well as the MOD17A2 GPP products (GPPMOD17). Based on validation with flux tower derived GPP estimates the results show that GPPNDVI3g is more accurate than GPPNDVI1g and is comparable in accuracy with GPPMOD15. In addition, GPPNDVI3g and GPPMOD15 have good spatial-temporal consistency. Our results indicate that GIMMS NDVI3g is an effective dataset for regional GPP simulation in Southeast Asia, capable of accurately tracking the variation and trends in long-term terrestrial ecosystem GPP dynamics.

Local Disparity Estimation With Three-Moded Cross Census and Advanced Support Weight

The classical local disparity methods use simple and efficient structure to reduce the computation complexity. To increase the accuracy of the disparity map, new local methods utilize additional processing steps such as iteration, segmentation, calibration and propagation, similar to global methods. In this paper, we present an efficient one-pass local method with no iteration. The proposed method is also extended to video disparity estimation by using motion information as well as imposing spatial temporal consistency. In local method, the accuracy of stereo matching depends on precise similarity measure and proper support window. For the accuracy of similarity measure, we propose a novel three-moded cross census transform with a noise buffer, which increases the robustness to image noise in flat areas. The proposed similarity measure can be used in the same form in both stereo images and videos. We further improve the reliability of the aggregation by adopting the advanced support weight and incorporating motion flow to achieve better depth map near moving edges in video scene. The experimental results show that the proposed method is the best performing local method on the Middlebury stereo benchmark test and outperforms the other state-of-the-art methods on video disparity evaluation.

Multisensor video fusion based on spatial–temporal salience detection

With three dimensional uniform discrete curvelet transform (3D-UDCT) and spatial–temporal structure tensor, a novel video fusion algorithm for videos with static background images is proposed in this paper. Firstly, the 3D-UDCT is employed to decompose source videos into many subbands with different scales and directions. Secondly, corresponding subbands of source videos are merged with different fusion schemes. Finally, the fused video is obtained by the reverse 3D-UDCT. Especially, when bandpass directional subband coefficients are merged, a spatial–temporal salience detection algorithm based on the structure tensor is performed. And each subband is divided into three types of regions, i.e., regions with temporal moving targets, regions with spatial features of background images, and smooth regions. Then different fusion rules are designed for each type of regions. Compared with some existing fusion methods, the proposed fusion algorithm can not only extract more spatial–temporal salient features from input videos but also perform better in spatial–temporal consistency. In addition, the proposed fusion algorithm can also be extended to fuse videos with dynamic background images by a simple modification. Several sets of experimental results demonstrate the feasibility and validity of the proposed fusion method.

Exploring Visual and Motion Saliency for Automatic Video Object Extraction

This paper presents a saliency-based video object extraction (VOE) framework. The proposed framework aims to automatically extract foreground objects of interest without any user interaction or the use of any training data (i.e., not limited to any particular type of object). To separate foreground and background regions within and across video frames, the proposed method utilizes visual and motion saliency information extracted from the input video. A conditional random field is applied to effectively combine the saliency induced features, which allows us to deal with unknown pose and scale variations of the foreground object (and its articulated parts). Based on the ability to preserve both spatial continuity and temporal consistency in the proposed VOE framework, experiments on a variety of videos verify that our method is able to produce quantitatively and qualitatively satisfactory VOE results.

Millennium-scale records of benthic foraminiferal communities from the central Great Barrier Reef reveal spatial differences and temporal consistency

Understanding long-term community dynamics and the ways in which they respond to major disturbances is a central management theme within coastal marine ecosystems. River outputs from the Queensland coastline directly affect inshore marine communities of the Great Barrier Reef (GBR), Australia. Of these, the Burdekin River exports the highest volume of terrestrial runoff. Following European settlement in the mid-19th century, over three quarters of the native vegetation from the Burdekin catchment were cleared for agricultural purposes. Despite such extensive historical catchment modification, the impact of these changes on the inshore GBR is largely unknown, primarily due to the paucity of long-term ecological data. To assess the effects of modern land-use change on inshore reef environments and to establish an historical baseline of community structure, we examined the sedimentary geochemistry and benthic foraminiferal assemblages of eight sediment cores collected from two coral reefs situated inside (Pandora) and outside (Havannah) an inner-shelf sediment prism formed during the Holocene. Foraminiferal community structure was reconstructed from the past millennium, and the time series was constrained using U-series dating of coral fragments within the cores. Environmental records were reconstructed using stable carbon isotopes (δ13C) and elemental C:N ratios from bulk sediment samples. Non-parametric analysis of community structure in benthic foraminifers indicated no change in community structure through time at either reef. Despite this apparent ecological persistence through time, significant differences in foraminiferal community structure were observed between the two reefs. The communities were clearly characterized by different functional groups; heterotrophic genera were persistent within, and symbiont-bearing genera were persistent outside, the Holocene inshore sediment wedge. We found no difference in the source of organic matter (interpreted from δ13C values) either between reefs or through time, yet elemental C:N ratios indicated a difference in the amount of organic matter between reefs. The influence of the Holocene inshore sediment wedge was demonstrated by the dissimilarity in sedimentary C:N ratios between the two reefs.

Land subsidence in the Nile Delta of Egypt observed by persistent scatterer interferometry

Land subsidence is a common problem in vulnerable deltas. The Nile Delta is no exception. The impacts of land subsidence are heightened by the economic, social and historical importance of the delta to Egypt. A major debate has evolved in the past two decades concerning whether the land surface of the Nile Delta is subsiding. The debate is certainly problematic in light of the fact that current measures of subsidence across the delta are rough estimates at best. To date, knowledge of subsidence rates in the delta is limited to long-term geologic averages that assume spatial uniformity and temporal consistency. In this study, we apply persistent scatterer interferometry (PSI) to measure the magnitude and monitor the spatial and temporal variations of land subsidence in the Nile Delta, during 1993–2000, using synthetic aperture radar interferometric data of 5.66 cm wavelength. The average measured rates of local subsidence in two major cities in the delta, namely Mansura and Greater Mahala, are –9 and –5 mm year–1, respectively. The observed deformation features imply that subsidence in both cities is controlled mainly by local groundwater processes. Our PSI measurements indicate that no regional subsidence has occurred in either city between 1993 and 2000. The slight regional subsidence that is expected to occur over time due to the natural compaction of deltaic sediments most likely has been masked by surface displacements caused by seasonal oscillations in the groundwater level.

Spatio-Temporally Consistent Novel View Synthesis Algorithm From Video-Plus-Depth Sequences for Autostereoscopic Displays

In this paper, we propose a novel algorithm to generate multiple virtual views from a video-plus-depth sequence for modern autostereoscopic displays. To synthesize realistic content in the disocclusion regions at the virtual views is the main challenging problem for this task. Spatial coherence and temporal consistency are the two key factors to produce perceptually satisfactory virtual images. The proposed algorithm employs the spatio-temporal consistency constraint to handle the uncertain pixels in the disocclusion regions. On the one hand, regarding the spatial coherence, we incorporate the intensity gradient strength with the depth information to determine the filling priority for inpainting the disocclusion regions, so that the continuity of image structures can be preserved. On the other hand, the temporal consistency is enforced by estimating the intensities in the disocclusion regions across the adjacent frames with an optimization process. We propose an iterative re-weighted framework to jointly consider intensity and depth consistency in the adjacent frames, which not only imposes temporal consistency but also reduces noise disturbance. Finally, for accelerating the multi-view synthesis process, we apply the proposed view synthesis algorithm to generate the intensity and depth maps at the leftmost and rightmost viewpoints, so that the intermediate views are efficiently interpolated through image warping according to the associated depth maps between the two synthesized images and their corresponding symmetric depths. In the experimental validation, we perform quantitative evaluation on synthetic data as well as subjective assessment on real video data with comparison to some representative methods to demonstrate the superior performance of the proposed algorithm.

Implementation of a Semiphysical Model for Examining Solar Radiation in the Midwest

A semiphysical solar radiation (SR) model is implemented to generate a new historical daily SR database for 53 locations in nine Midwestern and six adjacent states (available from the Midwestern Climate Center). This model estimates daily SR using standard hourly meteorological observations (surface atmospheric pressure and dewpoint temperature; cloud height and fractional sky cover by layer) as well as time of day, day of year, latitude/longitude, and the daily presence/absence of snow cover as input. Because of an extensive effort to interpolate for missing input (especially cloud) data, the daily SR dataset generated is 92% complete for all 53 stations for 1948–91, and 99% complete for the 43 stations with continuous hourly meteorological observations that commenced during 1945–50 and extended through 1991. Consistent with previous work, the model validates favorably against sets of daily SR measurements from (three) contrasting parts of the study region, and so its output is used here without adjustment. Analyses of the dataset document the basic Midwestern spatial and temporal SR variability since the mid-to late 1940s. The spatial variation of calendar monthly mean SR is dominated by a near-meridional (north-eastward) decrease in fall and winter. This fundamental pattern is substantially perturbed from midspring through summer by subregional-to-mesoscale variability around and across the Great Lakes. Time series of individual monthly station mean SR values exhibit a pronounced, regionwide 1945–91 downtrend for August–November. This decline is strongest (∼12%) and most statistically significant (>99% level) for October in a belt extending east-southeastward from west-central Wisconsin across southern lake Michigan and western Lake Erie to western Pennsylvania. The SR trends for December–July are largely positive but of lesser spatial coherence, temporal consistency, and statistical significance.