Daily MODIS 500 m reflectance anisotropy direct broadcast (DB) products for monitoring vegetation phenology dynamics

Land surface vegetation phenology is an efficient bio-indicator for monitoring ecosystem variation in response to changes in climatic factors. The primary objective of the current article is to examine the utility of the daily MODIS 500 m reflectance anisotropy direct broadcast (DB) product for monitoring the evolution of vegetation phenological trends over selected crop, orchard, and forest regions. Although numerous model-fitted satellite data have been widely used to assess the spatio-temporal distribution of land surface phenological patterns to understand phenological process and phenomena, current efforts to investigate the details of phenological trends, especially for natural phenological variations that occur on short time scales, are less well served by remote sensing challenges and lack of anisotropy correction in satellite data sources. The daily MODIS 500 m reflectance anisotropy product is employed to retrieve daily vegetation indices (VI) of a 1 year period for an almond orchard in California and for a winter wheat field in northeast China, as well as a 2 year period for a deciduous forest region in New Hampshire, USA. Compared with the ground records from these regions, the VI trajectories derived from the cloud-free and atmospherically corrected MODIS Nadir BRDF (bidirectional reflectance distribution function) adjusted reflectance (NBAR) capture not only the detailed footprint and principal attributes of the phenological events (such as flowering and blooming) but also the substantial inter-annual variability. This study demonstrates the utility of the daily 500 m MODIS reflectance anisotropy DB product to provide daily VI for monitoring and detecting changes of the natural vegetation phenology as exemplified by study regions comprising winter wheat, almond trees, and deciduous forest.     

Classification of new-ice in the Greenland Sea using Satellite SSM/I radiometer and SeaWinds scatterometer data and comparison with ice model

In the ice covered waters of the Greenland Sea the polarisation ratio of QuikSCAT SeaWinds Ku-band (13.4 GHz) scatterometer measurements and the polarisation ratio of DMSP-SSM/I 19 GHz radiometer measurements are used in combination to classify new-ice and mature ice. In particular, the formation of the new-(frazil/pancake)-ice ‘Odden’ (8° W, 75° N) March 11th-18th, 2001, is used in the study. The results of the ice cover classification in the Greenland Sea are compared to model parameters from a sea ice model. The classification of each ice pixel is performed using its backscatter and radiative properties as reflected in the polarisation ratio. Our results based on these comparisons show that the transformation into older mature (sheet) ice occurs within 5-10 days. During one day the new-ice cover increased by 33 000 km². The new-ice appears in March 2001 as a peninsula (maximum extent 56 000 km²) appended to the belt of older ice drifting along the East Coast of Greenland. These results are consistent with the ice model and with Radarsat images. Furthermore, using the ice model it is demonstrated that the new-ice/mature ice threshold in the classification corresponds to the physical transition of the ice cover from pancake ice to a consolidated young-ice sheet. The classification of each pixel into ice or water is done using two scatterometer parameters, namely the polarisation ratio and the daily standard deviation of the backscatter.     

Extracting grassland vegetation phenology in North China based on cumulative SPOT-VEGETATION NDVI data

Plant phenology is one of the main indicators of climate or other environmental processes. This paper assesses the detection accuracy of start of season (SOS) and end of season (EOS) for grassland vegetation in north China from 2001 to 2010 using SPOT-VEGETATION normalized difference vegetation index (NDVI) data sets and in situ observations. The cumulative NDVI is calculated and fitted using a logistic model to identify phenological transition dates. The curvature of the fitted logistic models predicts phenological transition dates that correspond to the times at which the curvature in the yearly integrated NDVI exhibits local minimums or maximums. Validating with in situ observations, phenological dates are extracted from satellite time series data and are accurate to within 10 days. The spatial trends of SOS and EOS are very similar for 2001–2010. SOS mainly occurs from the day of year (DOY) 110 to DOY 170, and EOS occurs from DOY 240 to DOY 300. SOS displays a marked delay from south to north, while EOS gradually advances, indicating regional differences in climate and terrain. However, the effect of latitude and longitude on the average EOS of alpine grasslands is not significantly different, while SOS at low latitude and high longitude is 10 days earlier than at high-latitude and high-longitude regions. We detected an overall advance in SOS of 3.1 days over 10 years, and a 1.3-day delay in EOS. However, the amplitude is low (about 5 days) and the changes in most regions are not significant (close to zero). The results in this paper are concordant with many reported studies that explored the phenology of grasslands in North China, which is an important component of global grasslands science.     

Sensitivity of vegetation phenology detection to the temporal resolution of satellite data

Vegetation phenology derived from satellite data has increasingly received attention for applications in environmental monitoring and modelling. The accuracy of phenological estimates, however, is unknown at the regional and global level because field validation data are insufficient. To assess the accuracy of satellite‐derived phenology, this study investigates the sensitivity of phenology detection to both the temporal resolution of sampling and the number of consecutive missing values (usually representing cloud cover) in the time series of satellite data. To do this, time series of daily vegetation index data for various ecosystems are modelled and simulated using data from Moderate‐Resolution Imaging Spectroradiometer (MODIS) data. The annual temporal data are then fitted using piecewise logistic functions, which are employed to calculate curvature change rate for detecting phenological transition dates. The results show that vegetation phenology can be estimated with a high precision from time series with temporal resolutions of 6–16 days even if daily data contains some uncertainties. If the temporal resolution is no coarser than 16 days for time series sampled using an average composite, the absolute errors are less than 3 days. On the other hand, the phase shift of temporal sampling is shown to have limited impacts on phenology detection. However, the accuracy of phenology detection may be reduced greatly if missing values in the time series of 16‐day MODIS data occur around the onsets of phenological transition dates. Even so, the probability that the absolute error in phenological estimates is greater than 5 days is less than 4% when only one period is missing in the time series of 16‐day data during vegetation growing seasons; this probability increases to 20% if there are two consecutive missing values.     

The performance of vegetation indices for operational monitoring of CORINE vegetation types

Vegetation monitoring has been performed using remotely sensed images to secure food production, prevent fires, and protect natural ecosystems. Recent satellite sensors, such as the Moderate Resolution Imaging Spectroradiometer (MODIS), provide frequent wide-scale coverage in multiple areas of the spectrum, allowing the estimation of a wide range of specialized vegetation indices (VIs), each offering several advantages. It is not, however, clear which VI performs better during operational monitoring of wide-scale vegetation patches, such as CORINE Land Cover (CLC) classes. The aim of this work was to investigate the performance of several VIs in operational monitoring of vegetation condition of CLC vegetation types, using Terra MODIS data. Comparison among the VIs within each CLC class was conducted using the sensitivity ratio, a statistical measure that has not been used to compare VIs and does not require calibration curves between each VI and a biophysical parameter. In addition, the VI’s sensitivity to factors such as the aspect, viewing angle, signal saturation, and partial cloud cover was estimated with correlation analysis in order to identify their operational monitoring ability. Results indicate the enhanced vegetation index as superior for monitoring vegetation condition among CLC types, but not always optimum in performance tests for operational monitoring.     

Mapping of Sahelian vegetation parameters from ERS scatterometer data with an evolution strategies algorithm

The West African Sahel rainfall regime is known for its spatio-temporal variability at different scales which has a strong impact on vegetation development. This study presents results of the combined use of a simple water balance model, a radiative transfer model and ERS scatterometer data to produce map of vegetation biomass and thus vegetation cover at a spatial resolution of 25 km. The backscattering coefficient measured by spaceborne wind scatterometers over Sahel shows a marked seasonality linked to the drastic changes of both soil and vegetation dielectric properties associated to the alternating dry and wet seasons. For lack of a direct observation, METEOSAT rainfall estimates are used to calculate temporal series of soil moisture with the help of a water balance model. This a priori information is used as input of the radiative transfer model that simulates the interaction between the radar wave and the surface components (soil and vegetation). Then, an inversion algorithm is applied to retrieve vegetation aerial mass from the ERS scatterometer data. Because of the nonlinear feature of the inverse problem to be solved, the inversion is performed using a global stochastic nonlinear inversion method. A good agreement is obtained between the inverse solutions and independent field measurements with mean and standard deviation of −54 and 130 kg of dry matter by hectare (kg DM/ha), respectively. The algorithm is then applied to a 350,000 km² area including the Malian Gourma and Seno region and a Sahelian part of Burkina Faso during two contrasted seasons (1999 and 2000). At the considered resolution, the obtained herbaceous mass maps show a global qualitative consistency (r²=0.71) with NDVI images acquired by the VEGETATION instrument.     

Assessment of the offshore wind resource in Japan with the ASCAT microwave scatterometer

We analysed wind speed and direction off the coast of Japan using data from the satellite-borne Advanced Scatterometer (ASCAT) and the Weather Research and Forecasting model (WRF), validated these data using in situ wind measurements from 20 buoys, and evaluated the effect of the long time intervals from ASCAT observations on wind resource assessment. More than 25 km from the coast, and at heights of 10 m, the ASCAT wind speed has negative biases of up to 3.4% and root mean square errors of up to 18.5%; its wind direction has 11° to 27° of mean absolute error compared to buoy measurements at a height of 10 m. These accuracies are better than either the expected accuracies reported in the technical manual or those simulated with WRF with its spatial resolution of 10 km. We also evaluated long-term average ASCAT wind speeds in comparison to 4- and 5-year averages of in situ buoy wind speeds measured at three buoys, with resulting differences of –0.3%, –6.3%, and – 1.6%. Furthermore, wind roses show that appearance frequencies of the ASCAT wind direction for the long term are in a good agreement with those of the measurements at the three buoys. Our results show that the ASCAT-derived wind speed and direction are appropriate more than 25 km from the coast, and that the long time interval between ASCAT observations has an insignificant effect on wind resource assessment, if at least 4 or 5 years of averaged ASCAT data are used.     

Assessment of JERS-1 SAR for monitoring secondary vegetation in Amazonia: II. Spatial,temporal, and radiometric considerations for operational monitoring

While the role of Synthetic Aperture Radar (SAR) in operational tropical forest monitoring has yet to be defined, it is nevertheless a critical technology for improving our understanding of deforestation and secondary vegetation in the tropics. In order to understand the role of this technology in operational monitoring a systematic evaluation, relative to other existing technologies of its performance is required. In this paper we evaluate the spatial, temporal, and noise constraints of JERS SAR data for mapping and monitoring biomass of secondary vegetation in Rondonia, Brazil. Our results indicate that the variability in stand estimates of biomass is high and that the source of the majority of the variability is not from speckle and the intrinsic texture of the secondary vegetation but likely due to differences in environmental conditions resulting in differential background scattering properties. Multitemporal analysis significantly improves biomass estimates to the point where it is possible to map changes in biomass. Slight reductions in the variability in estimates of normalized radar cross-section greatly improve biomass estimation.     

Interfacing NOAA/ANHRR NDVI and soil truth maps for monitoring vegetation phenology at a local scale in a heterogeneous landscape of Southern Italy

This paper discusses the preprocessing, clustering, and labelling steps of data supplied from NOAA Advanced Very High Radiometers (AVHRR) to monitor vegetation phenology in a complex area (Vulture Basin, Italy). Time cluster maps of Normalized Difference Vegetation Index (NDVI) are compared with a land use map and a Digital Elevation Model of the region. This study results show that AVHRR/NDVI well discriminates forested areas whatever the altitude may be; whereas the phenology of cultivated fields must be distinguished between plain and mountain phenology. The pixels not fitting into this picture mostly account for three peculiar microclimatic situations (two long and narrow valleys and a smooth, sunny mountain area).     

Satellite passive microwave remote sensing for monitoring global land surface phenology

Vegetation phenology characterizes seasonal life-cycle events that influence the carbon cycle and land-atmosphere water and energy exchange. We analyzed global phenology cycles over a six year record (2003-2008) using satellite passive microwave remote sensing based Vegetation Optical Depth (VOD) retrievals derived from daily time series brightness temperature (T_b) measurements from the Advanced Microwave Scanning Radiometer on EOS (AMSR-E) and other ancillary data inputs. The VOD parameter derives vegetation canopy attenuation at a given microwave frequency (18.7 GHz) and varies with canopy height, density, structure and water content. An error sensitivity analysis indicates that the retrieval algorithm can resolve the VOD seasonal cycle over a majority of global vegetated land areas. The VOD results corresponded favorably (p < 0.01) with vegetation indices (VIs) and leaf area index (LAI) information from satellite optical-infrared (MODIS) remote sensing, and phenology cycles determined from a simple bioclimatic growing season index (GSI) for over 82% of the global domain. Lower biomass land cover classes (e.g. savannas) show the highest correlations (R = 0.66), with reduced correspondence at higher biomass levels (0.03 < R < 0.51) and higher correlations for homogeneous land cover areas (0.41 < R < 0.83). The VOD results display a unique end-of-season signal relative to VI and LAI series, and may reflect microwave sensitivity to the timing of vegetation biomass depletion (e.g. leaf abscission) and associated changes in canopy water content (e.g. dormancy preparation). The VOD parameter is independent of and synergistic with optical-infrared remote sensing based vegetation metrics, and contributes to a more comprehensive view of land surface phenology.     

Atmospheric conditions for monitoring the long-term vegetation dynamics in the Amazon using normalized difference vegetation index

This study examined the effect of biomass-burning aerosols and clouds on the temporal dynamics of the normalized difference vegetation index (NDVI) exhibited by two widely used, time-series NDVI data products: the Pathfinder AVHRR land (PAL) dataset and the NASA Global Inventory Monitoring and Modeling Studies (GIMMS) dataset. The PAL data are 10-day maximum-value NDVI composites from 1982 to 1999 with corrections for Rayleigh scattering and ozone absorption. The GIMMS data are 15-day maximum-value NDVI composites from 1982 to 1999. In our analysis, monthly maximum-value NDVI was extracted from these datasets. The effects were quantified by comparing time-series of NDVI from PAL and GIMMS with observations from the SPOT/VEGETATION sensor and aerosol index data from the Total Ozone Mapping Spectrometer (TOMS), and results from radiative transfer simulation. Our analysis suggests that the substantial large-scale NDVI seasonality observed in the south and east Amazon forest region with PAL and GIMMS is primarily caused by variations in atmospheric conditions associated with biomass-burning aerosols and cloudiness. Reliable NDVI data can be typically acquired from April to July when such effects are relatively low, whereas there is a few effective NDVI data from September to December. In the central Amazon forest region, where aerosol loads are relatively low throughout the year, large-scale NDVI seasonality results primarily from seasonal variations in cloud cover. Careful treatment of these aerosol and cloud effects is required when using NDVI from PAL and GIMMS (or other source) to determine large-scale seasonal and interannual dynamics of vegetation greenness and ecosystem-atmosphere CO₂ exchange in the Amazon region.     

Real-time monitoring and short-term forecasting of land surface phenology

Land surface phenology is an important process for real-time monitoring and short-term forecasting in diverse land management, health, and hydrologic modeling applications. Yet current efforts to characterize phenological processes are limited by remote sensing challenges and lack of uncertainty estimates. Here, for a global distribution of phenologically and climatically similar phenoregions, we used the Advanced Very High Resolution Radiometer to develop a conceptually and computationally simple technique for real-time and forecast applications. Our overall approach was to analyze the phenological behavior of groups of pixels without recourse to smoothing or fitting. We used a 3-step initial process: (1) define a phenoregion specific normalized difference vegetation index threshold; (2) for all days from 1982-2003, calculate the percent of pixels above the threshold (PAT); (3) calculate daily 1982-2003 empirical distributions of PAT. For real-time monitoring, the current PAT may then be compared to the historical range of variability and visualized in relation to user-defined levels. Using similar concepts, we projected daily PAT up to one month in the future and compared predicted and actual dates at which a hypothetical PAT was reached. We found that the maximum lead-time of phenological forecasts could be analytically defined for user-specified uncertainty levels. The approach is adaptable to different remote sensing technologies and provides a foundation for ascribing a sequence of ground conditions (e.g. snowmelt, vegetative growth, pollen production, insect phenology) to remotely sensed land surface phenology observations.     

Effect of climate change on vegetation phenology of different land-cover types on the Tibetan Plateau

Research in vegetation phenology change has been one heated topic of current ecological and climate change study. The Tibetan Plateau, as the highest plateau of the earth, is more vulnerable and sensitive to climate change than many other regions. In this region, shifts in vegetation phenology have been intensively studied during recent decades, primarily based on satellite-retrieved data. In this study, we explored the spatiotemporal changes of vegetation phenology for different land-cover types in the Tibetan Plateau and characterized their relationship with temperature and precipitation by using long-term time-series datasets of normalized difference vegetation index (NDVI) from 1982 to 2014. Diverse phenological changes were observed for different land-cover types, with an advancing start of growing season (SOS), delaying end of growing season (EOS) and increasing length of growing season (LOS) in the eastern Tibetan Plateau where meadow was the dominant vegetation type, but with the opposite changes in the steppe and sparse herbaceous or sparse shrub regions which are mostly located in the northwestern and western edges of the Tibetan Plateau. Correlation analysis indicated that sufficient preseason precipitation may delay the SOS of evergreen forests in the southeastern Plateau and advance the SOS of steppe and sparse herbaceous or sparse shrub in relatively arid areas, while the advance of SOS in meadow areas could be related to higher preseason temperature. For EOS, because it is less sensitive to climate change than SOS, the response of EOS for different land-cover types to precipitation and temperature were more complicated across the Tibetan Plateau.     

Retrieval of bare soil and vegetation parameters from wind scatterometer measurements over three different climatic regions

Medium to low resolution (1-50 km) active microwave sensors such as spaceborne scatterometers and wide-swath mode synthetic aperture radars have great potential as tools for long term monitoring over land and ice. To optimise the use of this kind of data, the heterogeneity of the target and its effects on the radar measurements need to be investigated and modelled, particularly in the view of retrieving geophysical parameters. In this paper, wind scatterometer measurements over three different test sites, the NOPEX region in Sweden, the HAPEX-Sahel site in Niger and the Niger delta area in Nigeria, are analysed. For these regions, a forward model is developed by considering the backscatter contributions of the bare surface, the seasonal and evergreen vegetation and the open water areas. Colocated high spatial resolution SAR data and ground information are used to characterise the target scene. The model is then inverted to retrieve monthly soil roughness, dielectric properties and vegetation parameters. It is shown that the measurements contain enough information to characterise these three different regions and to monitor their temporal evolution. The retrieved values obtained for the bare surface and the vegetation parameters are consistent with ground measurements collected in these areas. Further improvements are achieved by incorporating the time scale variability of the variables investigated into the retrieval scheme.     

Red and photographic infrared linear combinations for monitoring vegetation

In situ collected spectrometer data were used to evaluate and quantify the relationships between various linear combinations of red and photographic infrared radiances and experimental plot biomass, leaf water content, and chlorophyll content. The radiance variables evaluated included the red and photographic infrared (IR) radiance and the linear combinations of the IR/red ratio, the square root of the IR/red ratio, the IR-red difference, the vegetation index, and the transformed vegetation index. In addition, the corresponding green and red linear combinations were evaluated for comparative purposes. Three data sets were used from June, September, and October sampling periods.Regression analysis showed the increased utility of the IR and red linear combinations vis-à-vis the same green and red linear combinations. The red and IR linear combinations had 7% and 14% greater regression significance than the green and red linear combinations for the June and September sampling periods, respectively.The vegetation index, transformed vegetation index, and square root of the IR/red ratio were the most significant, followed closely by the IR/red ratio. Less than a 6% difference separated the highest and lowest of these four ER and red linear combinations. The use of these linear combinations was shown to be sensitive primarily to the green leaf area or green leaf biomass. As such, these linear combinations of the red and photographic IR radiances can be employed to monitor the photosynthetically active biomass of plant canopies.     

Comparison of vegetation phenology in the western USA determined from reflected GPS microwave signals and NDVI

The normalized microwave reflection index (NMRI) is a measure of multipath scattering calculated daily from continuously operating GPS sites. GPS satellites transmit L-band microwave signals, and thus NMRI is sensitive to the amount of water in vegetation, not plant greenness or dry biomass. The sensing footprint is approximately 1000 m², although reflections from a distance of hundreds of metres are important at some sites. NMRI exhibits clear seasonal variations that are linked to the changes in vegetation water content that accompany plant growth and senescence. In this paper, NMRI and the normalized difference vegetation index (NDVI) are compared for the period 2008–2012. NMRI data are derived from 184 GPS sites in the western USA. NDVI data are from the 250 m, 16-day pixel containing each GPS station. Amplitude of the annual growth cycle and correlation between NMRI and NDVI are estimated, with and without lags. Phenology metrics are calculated from both indices (i.e. the start of the growing season, timing of peak growth, and season length).

NMRI and NDVI are correlated at most sites, but the degree of correlation varies regionally. Correlation is lowest in California and coastal regions (R = 0.25), where NDVI increases earlier in the spring than NMRI. It is highest for mountain and prairie sites (R = 0.66 and 0.73, respectively). Allowing for a lag between NMRI and NDVI greatly increases the correlation. The lag that yields the greatest correlation is nearly 30 days for the California sites (R = 0.71 with lag), but only 10 days for mountain and prairie sites (R = 0.78 and 0.85 with lag). There are clear differences in phenology metrics extracted from NMRI and NDVI that are consistent with the correlation-lag analysis. Using NMRI, there is a later start to the season, later peak day of the year, and shorter season length. The greatest differences are in California where NDVI start of the season is nearly 60 days earlier than that calculated from NMRI. These data suggest that green-up precedes increases in vegetation water content, with the duration of offset varying regionally. This study is the first to compare GPS-derived microwave reflectance data with NDVI at multiple sites, and highlights both opportunities and limitations offered by NMRI data.     

Potential of the remotely-derived products in monitoring ecosystem water use efficiency across grasslands in Northern China

Water use efficiency (WUE) has been recognized as a crucial parameter to describe the interrelationship between carbon and water cycling. Quantitative assessment to spatiotemporal dynamics of ecosystem WUE in grasslands is of vital importance, given the large proportion of grasslands on the Earth’s land surface. Through continuous eddy covariance (EC) measurements at seven grassland sites in Northern China, this study examined the seasonal and interannual variations of gross primary production (GPP), evapotranspiration (ET) and WUE across four typical grassland ecosystems along a water availability gradient. The highest WUE occurred at the alpine meadow ecosystem with 1.45 g C kg^?1 H₂O, followed by the temperate meadow steppe and the typical steppe. The desert steppe had the lowest WUE with 0.53 g C kg^?1 H₂O. In addition, the remotely-derived WUE estimates from the Moderate Resolution Imaging Spectroradiometer (MODIS) and the Breathing Earth System Simulator (BESS) GPP and ET products, were used to characterize the variability in tower-based WUE over these grassland ecosystems. Generally, WUE_BESS had more consistent seasonal trajectories with WUE_EC at most grassland sites compared to the variability of WUE_MOD. The highest square of Pearson correlation coefficient (R²) values of WUE_BESS were achieved in the alpine meadow sites (approximately 0.83), as well as the lowest root mean square error (RMSE), which ranged from 0.21 to 0.37 g C kg^?1 H₂O. However, the performances of both WUE_BESS and WUE_MOD lacked skills at the desert steppe sites which had low vegetation productivity. These remotely-derived WUE products, particularly the WUE_MOD, tended to overestimate the annual mean WUE_EC across these grassland types, with exception of the alpine meadow sites, where exhibited good performance. The underlying reasons for the biases of the MODIS- and BESS-based products in capturing the seasonal dynamics of grassland WUE were also examined. In general, GPP_MOD performed better than GPP_BESS over an 8-day period, whereas ET_BESS had a higher accuracy compared to ET_MOD across the different grassland ecosystems. Our analyses may be useful for improving the remote sensing-based GPP and ET products to accurately monitor the ecosystem WUE patterns of grasslands over large areas.     

Daily MODIS products for analyzing early season vegetation dynamics across the North Slope of Alaska

Monitoring the growth and distribution of Arctic tundra vegetation is important for understanding changes in early growing season conditions in Arctic ecosystems in response to a warming climate. The primary objective of this study is to examine the utility of computed Daily Moderate Resolution Imaging Spectroradiometer (MODIS) Normalized Difference Vegetation Index (NDVI) products relative to 16-day maximum value composite (MVC) datasets for observing early season green-up dynamics of Arctic tundra vegetation across the North Slope of Alaska. Greening in the Arctic typically occurs shortly after snowmelt and can potentially be captured by using satellite observations that are available on a daily basis. Daily MODIS Snow Cover products were employed to retrieve dates of complete snowmelt (DOCS) for 2003-2005 for pixels that were cloud free at the time of complete snowmelt. Given the sparseness of cloud-free observations in both space and time, early season NDVI trajectories for cloud-free pixels were derived using daily MODIS data based on two approaches: a chronosequence (temporally continuous but aspatial) and a pixel trajectory (temporally discontinuous but spatial explicit) approach.On average during the three-year period, 12.5% of the North Slope region was cloud free at the time of complete snowmelt and a majority of these cloud-free pixels (65%) were associated with the Coastal Plain province. In contrast, the Foothills region was relatively less cloudy from the time following complete snowmelt until peak greenness (56%) than the Coastal Plain province (61%). As a result, vegetation communities that lie mostly in the Foothills province such as shrub tundra and moist acidic tundra classes had more cloud-free observations available to characterize NDVI trajectories using the pixel trajectory approach. Complete snowmelt in the North Slope generally occurred between day of year (DOY) 140 and 170 over the three years with areas covered by the shrub tundra vegetation community (Foothills province) experiencing snowmelt first in all three years with mean DOCS ranging from DOY 148 in 2004 to DOY 158 in 2003. For approximately two weeks following complete snowmelt (Phase I, a period of rapid NDVI increase), the Daily NDVI derived trajectories were substantially different from the MVC NDVI trajectories. Early season integrated NDVI (ESINDVI) values computed for Phase I were 7% higher using the Daily NDVI approaches relative to those derived from the MVC MODIS data for the North Slope region. Following this initial period, until peak greenness (Phase 2, a period of gradual NDVI increase), the Daily and MVC trajectories were similar in shape and magnitude. This study demonstrates the utility of the Daily MODIS Snow product for assessing cloud cover and snowmelt patterns and Daily MODIS NDVI data for observing and detecting sharp and rapid changes in early season vegetation phenology as seen during Phase I.     

Differences in monitoring vegetation dynamics between Moderate Resolution Imaging Spectroradiometer Collection 5 and Collection 6 vegetation index products on the Loess Plateau,China

Accurately monitoring vegetation dynamics on the Loess Plateau (LP) is critical for evaluating the benefits of ecological restoration projects. The Moderate Resolution Imaging Spectroradiometer (MODIS) vegetation index (VI) product has been a primary data source for monitoring vegetation dynamics. However, MODIS Collection 5 (C5) VI products are known to be affected by sensor degradation, which has been addressed in the newly released MODIS Collection 6 (C6) VI products. Herein, we compared the spatiotemporal differences in vegetation dynamics between the Terra MODIS C5 and C6 data products and among different annual value retrieval methods for the LP during 2001–2016. Our results indicated a lower magnitude but a greener trend in the normalized difference vegetation index (NDVI), and areas with significant greening (p < 0.050) were found to increase by about 13%–16% from C5 to C6, depending on the retrieval method. Regions with either no particular trend or a downward trend in vegetation derived from the Terra-C5 NDVI mostly showed significant increasing trends based on the Terra-C6 NDVI. Moreover, the different retrieval methods also exhibited differences in the evaluation of vegetation dynamics, with the largest differences in terms of both magnitude and trend being identified with the annual maximum value method. This highlighted a compelling need to choose suitable methods in different regions for the retrieval of annual VI values, in order to facilitate more robust and comparable conclusions. Additionally, discrepancies also existed in the response of vegetation to climate variations between the Terra-C5 and C6 products for all three annual VI retrieval methods. Our findings, based on multiple products and analysis methods, may lead to improved understanding of both vegetation dynamics and their linkage to climate variables. The results suggest that caution be utilized when using only MODIS Terra-C5 products to evaluate vegetation dynamics and calibrate ecosystem models.