Time Series Of Satellite Observations Research Articles

Demystifying LandTrendr and CCDC temporal segmentation

• We reviewed LandTrendr and CCDC temporal segmentation approaches. • These methods have distinct spectral and temporal assumptions and fitting approaches. • Key considerations include scalability, sensitivity, reproducibility and accuracy. • Conceptual and methodological comparisons support informed use and future development. Improved access to remotely sensed imagery and time series algorithms in combination with increased availability of cloud computing resources and platforms such as Google Earth Engine have significantly expanded the community of users processing and analyzing time series of satellite observations. Though individual time series analysis methods and their applications tend to be well-documented, comparisons of different approaches are beneficial to new users faced with the choice of different algorithms and parameterizations. We review two temporal segmentation approaches that have become increasingly prevalent in land cover mapping and monitoring: LandTrendr (Landsat-based detection of Trends in Disturbance and Recovery) and CCDC (Continuous Change Detection and Classification). We examine differences in the way these approaches use the temporal and spectral domains and compare model specifications and outputs. This review highlights previous work and applications, current limitations, ongoing challenges, and opportunities for future integration and comparison of methods and map products, and is expected to benefit both user and developer communities.

A Simple Approach for Mapping Forest Cover from Time Series of Satellite Data

Forest cover mapping based on multi-temporal satellite observations usually uses dozens of features as inputs, which requires huge training data and leads to many ill effects. In this paper, a simple but efficient approach was proposed to map forest cover from time series of satellite observations without using classifiers and training data. This method focuses on the key step of forest mapping, i.e., separation of forests from herbaceous vegetation, considering that the non-vegetated area can be easily identified by the annual maximum vegetation index. We found that the greenness of forests is generally stable during the maturity period, but a similar greenness plateau does not exist for herbaceous vegetation. It means that the mean greenness during the vegetation maturity period of forests should be larger than that of herbaceous vegetation, while its standard deviation should be smaller. A combination of these two features could identify forests with several thresholds. The proposed approach was demonstrated for mapping the extents of different forest types with MODIS observations. The results show that the overall accuracy ranges 91.92–95.34% and the Kappa coefficient is 0.84–0.91 when compared with the reference datasets generated from fine-resolution imagery of Google Earth. The proposed approach can greatly simplify the procedures of forest cover mapping.

Using Landsat Data to Assess the Status of Coral Reefs Cover along the Red Sea Coast, Egypt

The Egyptian Red Sea coast has experienced rapid development since the 1970’s. In particular, the coastal area near the City of Hurghada has been transformed into a long strip of touristic villages and hotels in this short time span. This is an area that has historically had abundant and diverse coral reef communities. To assess possible impacts on benthic coral reef cover in the region, Landsat satellite data collected over the Red Sea from 1973 through 2015 were analyzed to estimate urban expansion in the Hurghada region, shoreline changes, and changes in coral reef cover over time. A time series of satellite observations using Landsat 1-MSS, Landsat 5-TM, Landsat 7-ETM+, and Landsat 8-OLI was assembled, with images acquired in 1973, 1984, 1987, 1992, 1998, 2000, 2003, 2010, and 2015. Images were geometrically, radiometrically and atmospherically corrected, and a water-column correction was implemented prior to comparing images to assess change in landcover. The results show that during the last 42 years, the coral reef cover decreased 6.21 Km² while the built coastal area increased 13.4 Km². These observations were used to compute total economic value (TEV) of coral reef habitats and the cost of degradation in terms of physical losses of coral reef area which equals about18.63$ Billion.

Quantifying Surface Water Dynamics at 30 Meter Spatial Resolution in the North American High Northern Latitudes 1991–2011

The availability of a dense time series of satellite observations at moderate (30 m) spatial resolution is enabling unprecedented opportunities for understanding ecosystems around the world. A time series of data from Landsat was used to generate a series of three maps at decadal time step to show how surface water has changed from 1991 to 2011 in the high northern latitudes of North America. Previous attempts to characterize the change in surface water in this region have been limited in either spatial or temporal resolution, or both. This series of maps was generated for the NASA Arctic and Boreal Vulnerability Experiment (ABoVE), which began in fall 2015. These maps show a “nominal” extent of surface water by using multiple observations to make a single map for each time step. This increases the confidence that any detected changes are related to climate or ecosystem changes not simply caused by short duration weather events such as flood or drought. The methods and comparison to other contemporary maps of the region are presented here. Initial verification results indicate 96% producer accuracy and 54% user accuracy when compared to 2-m resolution WorldView-2 data. All water bodies that were omitted were one Landsat pixel or smaller, hence below detection limits of the instrument.

Ice-dammed lake drainage cools and raises surface salinities in a tidewater outlet glacier fjord, west Greenland

The drainage of ice-dammed lakes in the form of outburst floods in Greenland is detected regularly by remote sensing, and these events are expected to occur more frequently in a warmer climate. However, their impact on ice sheet stability and neighboring water bodies is still unknown. In this interdisciplinary study, we investigate lake drainages from the Greenland Ice Sheet into a west Greenland fjord by analyzing simultaneous time series of satellite observations and direct hydrographic measurements of temperature and salinity in the fjord. Satellite images show that, in general, lake drainages have occurred quasiperiodically during the last decade. A particular sequence of drainage events was observed by satellite in 2009 and was analyzed together with the first direct hydrographic observations. Signs of ice-dammed lake drainages were observed by a downstream mooring located just below the intertidal zone. The release of freshwater occurred at the fjord subsurface at a tidewater outlet glacier. The downstream in-water sequence of property changes in relation to these drainage events was observed as an almost immediate decrease in surface layer temperature (~2°C) followed within a week by the arrival of a high-saline pulse (~ +5 units) with elevated salinity lasting for several days during the passage. During lake drainages, large amounts of relatively warm and saline intermediate water are brought to the near-surface layers by entrainment processes near the glacier front, and this influences the hydrography of the fjord but also impacts the ecosystem through upwelling of nutrient-rich intermediate water.

Identification of environmental anomaly hot spots in West Africa from time series of NDVI and rainfall

Studies of the impact of human activity on vegetation dynamics of the Sahelian belt of Africa have been recently re-invigorated by new scientific findings that highlighted the primary role of climate in the drought crises of the 1970s–1980s. Time series of satellite observations revealed a re-greening of the Sahelian belt that indicates no noteworthy human effect on vegetation dynamics at sub continental scale from the 1980s to late 1990s. However, several regional/local crises related to natural resources occurred in the last decades despite the re-greening thus underlying that more detailed studies are needed. In this study we used time-series (1998–2010) of SPOT–VGT NDVI and FEWS–RFE rainfall estimates to analyse vegetation – rainfall correlation and to map areas of local environmental anomalies where significant vegetation variations (increase/decrease) are not fully explained by seasonal changes of rainfall. Some of these anomalous zones (hot spots) were further analysed with higher resolution images Landsat TM/ETM+ to evaluate the reliability of the identified anomalous behaviour and to provide an interpretation of some example hot spots. The frequency distribution of the hot spots among the land cover classes of the GlobCover map shows that increase in vegetation greenness is mainly located in the more humid southern part and close to inland water bodies where it is likely to be related to the expansion/intensification of irrigated agricultural activities. On the contrary, a decrease in vegetation greenness occurs mainly in the northern part (12°–15°N) in correspondence with herbaceous vegetation covers where pastoral and cropping practices are often critical due to low and very unpredictable rainfall. The results of this study show that even if a general positive re-greening due to increased rainfall is evident for the entire Sahel, some local anomalous hot spots exist and can be explained by human factors such as population growth whose level reaches the ecosystem carrying capacity as well as population displacement leading to vegetation recovery.

Quantitative analysis of the influence of dust sea surface forcing on the primary production of the subtropical Atlantic Ocean using a ten-year time series of satellite observations

[1] Dust aerosols that are not deposited over oceans are able to significantly reduce the solar energy available at the sea surface. Here, the impact of dust aerosols on the photosynthetically available radiation (PAR) at the sea surface and on the associated oceanic primary production (PP) is quantified over the subtropical Atlantic Ocean based on a ten-year time series of satellite observations. The ten-year average value of the attenuation of bothPAR and PP due to dust aerosols is high (∼15%). The comparisons with predictions suggested that the decrease of PP might be ∼35% in the case of intense episodic events (i.e., dust aerosols optical depth > 0.6). Therefore, dust aerosol events could significantly alter the organic carbon budget of the underlying oceanic ecosystems. The analysis of the interannual variations of the relative reduction of primary production (ΔPP/PP) due to dust aerosols showed that the evolution of ΔPP/PP does not exhibit any major trend of variation within the entire study area over the decade. However, a significant tendency (0.22% per year) is found near Africa in summer. Thus, dust aerosol events might induce a major decrease of the marine productivity the next centuries. The radiative forcing of dust aerosols on the sea surface needs to be accounted for in coupled atmosphere–ocean models for calculating correctly the primary production. A more extensive analysis of the aerosol radiative budget is also required to better understand the link between the atmospheric and oceanic processes driving the primary production over dust aerosol areas.

Post‐fire changes in net shortwave radiation along a latitudinal gradient in boreal North America

Understanding how a changing boreal fire regime is likely to influence regional climate requires detailed information about fire effects on the surface radiation budget. We used time series of satellite observations of surface albedo from 2000–2011 and fire perimeters since 1970 to study post‐fire changes in surface net shortwave radiation along a latitudinal transect in central Canada. Fire‐induced surface shortwave forcing (SSF) integrated over an annual cycle for the first 30 years after fire was similar (−4.1 W m−2 with a 95% confidence interval of −4.5 to −3.7 W m−2) between southern and northern boreal regions. The lack of a latitudinal difference in SSF was caused by counteracting latitudinal trends in seasonal contributions. Spring (March, April, and May) SSF increased with latitude, from −7.2 W m−2 in the south to −10.1 W m−2in the north, primarily because of delayed snow melt, which amplified albedo differences between unburned forests and recovering stands. In contrast, winter incoming solar radiation and summer albedo change decreased from south to north, resulting in a decreasing latitudinal trend in winter and summer SSF. Vegetation recovery was slower in the north, leading to smaller increases in summer albedo during the first decade after fire, and a prolonged phase of elevated spring albedo during the second decade. Our results indicate that fires reduce surface net shortwave radiation considerably for many boreal forest ecosystems in North America, providing further evidence that disturbance‐mediated shifts in surface energy exchange need to be considered in efforts to manage these forests for climate change mitigation.

Evaluation of a Global Vegetation Model using time series of satellite vegetation indices

Abstract. Atmospheric CO2 drives most of the greenhouse effect increase. One major uncertainty on the future rate of increase of CO2 in the atmosphere is the impact of the anticipated climate change on the vegetation. Dynamic Global Vegetation Models (DGVM) are used to address this question. ORCHIDEE is such a DGVM that has proven useful for climate change studies. However, there is no objective and methodological way to accurately assess each new available version on the global scale. In this paper, we submit a methodological evaluation of ORCHIDEE by correlating satellite-derived Vegetation Index time series against those of the modeled Fraction of absorbed Photosynthetically Active Radiation (FPAR). A perfect correlation between the two is not expected, however an improvement of the model should lead to an increase of the overall performance. We detail two case studies in which model improvements are demonstrated, using our methodology. In the first one, a new phenology version in ORCHIDEE is shown to bring a significant impact on the simulated annual cycles, in particular for C3 Grasses and C3 Crops. In the second case study, we compare the simulations when using two different weather fields to drive ORCHIDEE. The ERA-Interim forcing leads to a better description of the FPAR interannual anomalies than the simulation forced by a mixed CRU-NCEP dataset. This work shows that long time series of satellite observations, despite their uncertainties, can identify weaknesses in global vegetation models, a necessary first step to improving them.

The development of a Canadian dynamic habitat index using multi-temporal satellite estimates of canopy light absorbance

Monitoring patterns of fauna diversity across the landscape, both spatially and temporally, presents special challenges due to the dynamic nature of populations and complex interactions with the local and regional environment. One area where progress is being made is the development of relationships between regional biodiversity with indirect indicators or surrogates, such as vegetative production. In this paper we discuss implementation of a dynamic habitat index, originally developed in Australia, to Canadian conditions. The index, based on the fraction of photosynthetically active radiation (fPAR) absorbed by vegetation, a variable which is analogous to green vegetation cover, is derived solely from satellite data. The index utilizes time series of satellite observations of greenness to derive three indicators of the underlying vegetation dynamics; the cumulative annual greenness, the minimum level of perennial cover, and the degree of vegetation seasonality. We apply the index across Canada and compare the three components by ecozones, demonstrating that Canada's terrestrial environment can effectively be clustered into five major dynamic habitat regimes. These range from those with low cumulative greenness and highly seasonal variation in cover, to regimes which have high canopy light absorbance with limited seasonality and continuous annual green cover. By comparing data from multiple years, our analysis indicates that a number of these ecozones have experienced changes in their composition over the past 6 years. We believe this methodology can provide an initial stratification of large areas for biodiversity monitoring and can be used to focus finer scale approaches to specific regions of interest or monitor regions too remote for comprehensive field surveys.

Intraseasonal Convective Perturbations Related to the Seasonal March of the Indo-Pacific Monsoons

Abstract The seasonal evolution of the 20–90-day intraseasonal variability (ISV) of the convection in the Indo-Pacific region presents intriguing features related to monsoon dynamics: (i) a sharp ISV maximum in May for the southern Bay of Bengal and in June for the eastern Arabian Sea, (ii) a maximum ISV over the west Pacific in July–September when the ISV over the northern Indian Ocean is weaker, (iii) a persistent ISV north of Australia from December to March, and (iv) a weak ISV over continental regions. The source of these behaviors is investigated from time series of satellite observations, meteorological reanalysis, and from an ocean mixed layer depth (MLD) climatology. For the northern Indian Ocean, sharp ISV maxima in May and June, when the MLD is still small (∼20–30 m), are related to an abrupt surface cooling associated with the setting of the monsoon low-level jet. The ISV of the convection is smaller over these regions during July–September when the MLD is deepened (∼60–70 m) by the monsoon low-level wind forcing. At this time, the low-level wind is weak over the west Pacific, the MLD is small, and the amplitude of the ISV remains large. North of Australia, and also over the south equatorial Indian Ocean, there is no particular reinforcement of the ISV near the onset. A weak low-level jet and a small MLD prevail during the whole monsoon season in good agreement with a more uniform ISV. These results support the hypothesis that a moderately shallow (∼20–30 m) ocean mixed layer is important for maintaining locally the ISV of the convection, when the conditions (e.g., SST, large-scale circulation) for such an ISV to exist are satisfied. The seasonal variability of the MLD, related in part to the monsoon low-level jet variability over the tropical oceans, thus gives a link between the seasonal march of monsoon circulations and the seasonal evolution of the ISV of the convection. This has implications for a better understanding of the origin of the ISV and for the predictability of its amplitude, especially near the monsoon onset date.

Forecasting front displacements with a satellite based ocean forecasting (SOFT) system

Relatively long term time series of satellite data are nowadays available. These spatio–temporal time series of satellite observations can be employed to build empirical models, called satellite based ocean forecasting (SOFT) systems, to forecast certain aspects of future ocean states. The forecast skill of SOFT systems predicting the sea surface temperature (SST) at sub-basin spatial scale (from hundreds to thousand kilometres), has been extensively explored in previous works. Thus, these works were mostly focussed on predicting large scale patterns spatially stationary. At spatial scales smaller than sub-basin (from tens to hundred kilometres), spatio–temporal variability is more complex and propagating structures are frequently present. In this case, traditional SOFT systems based on Empirical Orthogonal Function (EOF) decompositions could not be optimal prediction systems. Instead, SOFT systems based on Complex Empirical Orthogonal Functions (CEOFs) are, a priori, better candidates to resolve these cases. In this work we study and compare the performance of an EOF and CEOF based SOFT systems forecasting the SST at weekly time scales of a propagating mesoscale structure. The SOFT system was implemented in an area of the Northern Balearic Sea (Western Mediterranean Sea) where a moving frontal structure is recurrently observed. Predictions from both SOFT systems are compared with observations and with the predictions obtained from persistence models. Results indicate that the implemented SOFT systems are superior in terms of predictability to persistence. No substantial differences have been found between the EOF and CEOF-SOFT systems.

Satellite Monitoring of Vegetation Phenology and Fire Fuel Conditions in Hawaiian Drylands

Abstract Grass-fueled fires accelerate grassland expansion into dry Hawaiian woodlands by destroying native forests and by producing a disturbance regime that favors grass-dominated plant communities. Knowledge of grassland phenology is a key component of ecosystem assessments and fire management in Hawaii, but diverse topographic relief and poor field-sampling capabilities make ground studies impractical. Remote sensing offers the best approach for large-scale, spatially contiguous measurements of dryland vegetation phenology and fire fuel conditions. A 500-m spatial resolution, 8-day temporal resolution Terra Moderate Resolution Imaging Spectroradiometer (MODIS) satellite time series of photosynthetic vegetation (PV), nonphotosynthetic vegetation (NPV), and exposed substrate conditions was developed for the island of Hawaii between 2000 and 2004. The results compared favorably with similar measurements of drylands from higher-resolution aircraft data. The satellite time series was compared with available environmental data on precipitation, fire history, and grazing intensity. From these analyses, the temporal patterns of PV and its conversion to NPV and finally to bare substrate were observed. An NPV buildup following fire of 7–8 yr was projected, and more heavily grazed lands were found to exhibit reduced NPV cover, most notably during the summer fire season. These results demonstrate the effects that land use and disturbance history have on fire conditions, and they support the concept that grazed lands managed to reduce litter buildup pose a lower risk of fire across ample geographic scales. Time series of satellite observations with modern analysis techniques can be used with environmental data to support a regional fire-monitoring program throughout Hawaii.

Real‐time forecasting at weekly timescales of the SST and SLA of the Ligurian Sea with a satellite‐based ocean forecasting (SOFT) system

Satellites are the only systems able to provide continuous information on the spatiotemporal variability of vast areas of the ocean. Relatively long‐term time series of satellite data are nowadays available. These spatiotemporal time series of satellite observations can be employed to build empirical models, called satellite‐based ocean forecasting (SOFT) systems, to forecast certain aspects of future ocean states. SOFT systems can predict satellite‐observed fields at different timescales. The forecast skill of SOFT systems forecasting the sea surface temperature (SST) at monthly timescales has been extensively explored in previous works. In this work we study the performance of two SOFT systems forecasting, respectively, the SST and sea level anomaly (SLA) at weekly timescales, that is, providing forecasts of the weekly averaged SST and SLA fields with 1 week in advance. The SOFT systems were implemented in the Ligurian Sea (Western Mediterranean Sea). Predictions from the SOFT systems are compared with observations and with the predictions obtained from persistence models. Results indicate that the SOFT system forecasting the SST field is always superior in terms of predictability to persistence. Minimum prediction errors in the SST are obtained during winter and spring seasons. On the other hand, the biggest differences between the performance of SOFT and persistence models are found during summer and autumn. These changes in the predictability are explained on the basis of the particular variability of the SST field in the Ligurian Sea. Concerning the SLA field, no improvements with respect to persistence have been found for the SOFT system forecasting the SLA field.

Satellite confirmation of the dominance of chlorofluorocarbons in the global stratospheric chlorine budget

OBSERVED increases in concentrations of chlorine in the stratosphere1–7 have been widely implicated in the depletion of lower-stratospheric ozone over the past two decades8–14. The present concentration of stratospheric chlorine is more than five times that expected from known natural 'background' emissions from the oceans and biomass burning15–18, and the balance has been estimated to be dominantly anthropogenic in origin, primarily due to the breakdown products of chlorofluorocarbons (CFCs)19,20. But despite the wealth of scientific data linking chlorofluorocarbon emissions to the observed chlorine increases, the political sensitivity of the ozone-depletion issue has generated a re-examination of the evidence21,22. Here we report a four-year global time series of satellite observations of hydrogen chloride (HC1) and hydrogen fluoride (HF) in the stratosphere, which shows conclusively that chlorofluorocarbon releases—rather than other anthropogenic or natural emissions—are responsible for the recent global increases in stratospheric chlorine concentrations. Moreover, all but a few per cent of observed stratospheric chlorine amounts can be accounted for by known natural and anthropogenic tropospheric emissions. Altogether, these results implicate the chlorofluorocarbons beyond reasonable doubt as dominating ozone depletion in the lower stratosphere.

A bidirectional reflectance model of the Earth's surface for the correction of remote sensing data

A surface bidirectional reflectance model has been developed for the correction of surface bidirectional effects in time series of satellite observations, where both sun and viewing angles are varying. The model follows a semiempirical approach and is designed to be applicable to heterogeneous surfaces. It contains only three adjustable parameters describing the surface and can potentially be included in an algorithm of processing and correction of a time series of remote sensing data. The model considers that the observed surface bidirectional reflectance is the sum of two main processes operating at a local scale: (1) a diffuse reflection component taking into account the geometrical structure of opaque reflectors on the surface, and shadowing effects, and (2) a volume scattering contribution by a collection of dispersed facets which simulates the volume scattering properties of canopies and bare soils. Detailed comparisons between the model and in situ observations show satisfactory agreement for most investigated surface types in the visible and near‐infrared spectral bands. The model appears therefore as a good candidate to reduce substantially the undesirable fluctuations related to surface bidirectional effects in remotely sensed multitemporal data sets.