Forecasting System Development for Nonlinear and Nonstationary Time Series of Normalized Difference Vegetation Index

MODerate-resolution Imaging Spectroradiometer (MODIS) vegetation index (VI) time-series data are among the most widely utilized remote sensing datasets. To improve the quality of MODIS VI time-series data, most prior methods have focused on correcting negatively biased VI noise by approaching the upper envelope of the VI time series. Such treatment, however, may cause overcorrections on some true local low VI values, resulting in inaccurate simulations of vegetation phenological characteristics. In addition, another challenge in reconstructing MODIS VI time series is to fill temporally continuous gaps. The earlier spatiotemporal Savitzky-Golay (STSG) method tackled this problem by utilizing multiyear VI data, but its performance heavily relies on the consistency of data across different years. In this study, we proposed an improved STSG (iSTSG) method. The new method accounts for the autocorrelation within the VI time series and fills missing values in the VI time series by leveraging spatiotemporal VI data from the current year alone. Furthermore, iSTSG incorporates an indicator to quantify potential overcorrections in the VI time series, aiming to more accurately simulate phenological characteristics. The experiments to reconstruct MODIS normalized difference VI (NDVI) time-series product (MOD13A2) at four typical sites (a million square kilometers for each site) suggest two clear advantages in iSTSG over the iterative SG (called Chen-SG) and STSG methods. First, iSTSG more accurately reconstructs the annual NDVI time series, exhibiting the smallest mean absolute differences (MADs) between the smoothed and the simulated reference NDVI time series (0.012, 0.018, and 0.020 for iSTSG, STSG, and Chen-SG, respectively). Second, iSTSG more effectively simulates phenological characteristics in the NDVI time series, including the onset dates for vegetation greenup and dormancy, as well as the crop harvest period. The advantages of iSTSG were also demonstrated when applied to the successor of MODIS, Visible Infrared Imaging Radiometer Suite (VIIRS) VI time-series product (VNP13A1). iSTSG can be implemented on the Google Earth Engine (GEE), offering significant benefits for various applications, particularly in crop mapping and vegetation/crop phenology studies.

HJ-1A/B is the first small satellite constellation built by China for environmental and disaster monitoring and forecasting. The satellite group has a 2-day repetition cycle and 30 m spatial resolution (charge coupled device camera). Thus, HJ-1A/B can provide hyper-temporal normalized difference vegetation index (NDVI) time series with a medium–high spatial resolution. However, the quality of the HJ NDVI time series can be abnormally low due to a number of factors, such as cloud cover, continuous fog, and haze. In the rainy season or in areas with serious atmospheric pollution, low-quality series often appear in succession, which is referred to as an abnormal segment. Neither the composition method nor quality flags satisfactorily solve this problem; therefore, a large amount of noise and long periods of abnormally low values often remain in HJ NDVI time series. This article presents a method to reconstruct the abnormal segments in HJ NDVI time series with the assistance of Moderate Resolution Imaging Spectroradiometer (MODIS) NDVI time series. The cointegration test was adopted to decide whether MODIS can be used for the reconstruction of NDVI time series for the corresponding HJ image pixels. Statistical quality control methods were used for singling out the abnormal segments in the HJ NDVI time series and establishing an error correction model that combines MODIS and HJ NDVI time series to perform the reconstruction. The study area is located in Jiangsu Province, China. Four-year (2009–2012) HJ multispectral images that cover the study area were used. The results show that abnormal segments in the HJ NDVI time series can be corrected using the proposed method. In a particular year, this method can decrease the root mean square error between the HJ NDVI time series and the reference sequence by 52.5%.

ABSTRACTChange detection within non-stationary and unequally spaced remote sensing time series has become a key methodology for a broad range of environmental applications. A new method of analysing vegetation variation over lands is proposed. Four regions in northern Tunisia with various characteristics are selected, and a non-stationary and unequally spaced Normalized Difference Vegetation Index (NDVI) time series is obtained for each region since 2000. The Landsat 7 remote sensing satellite imagery with insignificant cloud-shadow coverage is used to calculate the NDVI after atmospheric correction. The Least-Squares Wavelet (LSWAVE) software is implemented to rigorously analyse each NDVI time series and study the relationship between the vegetation of olive trees and temperature/precipitation in one of the regions. To investigate possible effects of temperature on the green cover caused by increasing water salinity, the coherency between the NDVI and sea surface temperature time series is also shown in the region of Lake Ichkeul in Tunisia.

Validation of synthetic daily Landsat NDVI time series data generated by the improved spatial and temporal data fusion approach

Normalized difference vegetation index (NDVI) time series data, derived from optical images, play a crucial role for crop mapping and growth monitoring. Nevertheless, optical images frequently exhibit spatial and temporal discontinuities due to cloudy and rainy weather conditions. Existing algorithms for reconstructing NDVI time series using multi-source remote sensing data still face several challenges. In this study, we proposed a novel method, an enhanced gap-filling and Whittaker smoothing (EGF-WS), to reconstruct NDVI time series (EGF-NDVI) using Google Earth Engine. In EGF-WS, NDVI calculated from MODIS, Landsat-8, and Sentinel-2 satellites were combined to generate high-resolution and continuous NDVI time series data. The MODIS NDVI was employed as reference data to fill missing pixels in the Sentinel–Landsat NDVI (SL-NDVI) using the gap-filling method. Subsequently, the filled NDVI was smoothed using a Whittaker smoothing filter to reduce residual noise in the SL-NDVI time series. With reference to the all-round performance assessment (APA) metrics, the performance of EGF-WS was compared with the conventional gap-filling and Savitzky–Golay filter approach (GF-SG) in Fusui County of Guangxi Zhuang Autonomous Region. The experimental results have demonstrated that the EGF-WS can capture more accurate spatial details compared with GF-SG. Moreover, EGF-NDVI of Fusui County exhibited a low root mean square error (RMSE) and a high coefficient of determination (R2). In conclusion, EGF-WS holds significant promise in providing NDVI time series images with a spatial resolution of 10 m and a temporal resolution of 8 days, thereby benefiting crop mapping, land use change monitoring, and various ecosystems, among other applications.

High quality Normalized Difference Vegetation Index (NDVI) time series data are important for regional and global ecological applications. However, missing observations in cloudy and shaded areas in current high-resolution satellite data severely limit their use. Traditional temporal filtering reconstruction methods are easily affected by frequent fluctuations, large data gaps and irregular noise in the observations. Meanwhile, satellite images often lack complete and accurate real data as training sets, which limits deep learning-based reconstruction methods. To address these problems, this paper proposes a reconstruction method with Landsat-8 NDVI time-series images based on bidirectional long-short memory networks (Bi-LSTM). The core of the method is a sequence-to-sequence time series reconstruction model using Bi-LSTM as an encoder, decoder, and additionally adding a mask layer to locate irregular gaps in the time series data. To train the model, a multi-scale continuous NDVI time series dataset was constructed using near-surface observations and satellite fused images. Simulation experiments were conducted in 48 areas covering PhenoCam sites in North America. The coefficient of determination (R2) between the simulation results of this method and the real observations reached 0.7024 with an absolute bias of -0.0080, which outperformed the filtering method both in predicting missing cloud-covered areas and in extracting key nodes for vegetation growth. This method is pre-trained on a large-scale ground sample dataset before fine-tuning the parameters for the scenarios, which can be easily and quickly applied to generate long-term continuous NDVI datasets at a large spatial scale. In addition, the analysis of the phenology monitoring in the experimental areas shows that the results extracted from the reconstructed results of this method is very close to the PhenoCam-measured values, which is significantly improved compared to the extraction results using original Landsat data only, and at the same time shows more spatial detail compared to the MODIS climate products. This indicates that the 30m gap-free NDVI time series is sufficient for analysis and understanding of climate dynamics in heterogeneous areas with multiple vegetation growth patterns, mixing plant components and microclimates variation, which are difficult to be identified by the coarse spatial resolution phenology product.

Timely and accurate crop type distribution maps are an important inputs for crop yield estimation and production forecasting as multi-temporal images can observe phenological differences among crops. Therefore, time series remote sensing data are essential for crop type mapping, and image composition has commonly been used to improve the quality of the image time series. However, the optimal composition period is unclear as long composition periods (such as compositions lasting half a year) are less informative and short composition periods lead to information redundancy and missing pixels. In this study, we initially acquired daily 30 m Normalized Difference Vegetation Index (NDVI) time series by fusing MODIS, Landsat, Gaofen and Huanjing (HJ) NDVI, and then composited the NDVI time series using four strategies (daily, 8-day, 16-day, and 32-day). We used Random Forest to identify crop types and evaluated the classification performances of the NDVI time series generated from four composition strategies in two studies regions from Xinjiang, China. Results indicated that crop classification performance improved as crop separabilities and classification accuracies increased, and classification uncertainties dropped in the green-up stage of the crops. When using daily NDVI time series, overall accuracies saturated at 113-day and 116-day in Bole and Luntai, and the saturated overall accuracies (OAs) were 86.13% and 91.89%, respectively. Cotton could be identified 40∼60 days and 35∼45 days earlier than the harvest in Bole and Luntai when using daily, 8-day and 16-day composition NDVI time series since both producer’s accuracies (PAs) and user’s accuracies (UAs) were higher than 85%. Among the four compositions, the daily NDVI time series generated the highest classification accuracies. Although the 8-day, 16-day and 32-day compositions had similar saturated overall accuracies (around 85% in Bole and 83% in Luntai), the 8-day and 16-day compositions achieved these accuracies around 155-day in Bole and 133-day in Luntai, which were earlier than the 32-day composition (170-day in both Bole and Luntai). Therefore, when the daily NDVI time series cannot be acquired, the 16-day composition is recommended in this study.

Recent studies have used satellite-derived normalized difference vegetation index (NDVI) time series to explore geographic patterns in boreal forest greening and browning. A number of these studies indicate that boreal forests are experiencing widespread browning, and have suggested that these patterns reflect decreases in forest productivity induced by climate change. Here we use NDVI time series from Landsat, which has much higher quality and spatial resolution than imagery used in most previous studies, to characterize biogeographic patterns in greening and browning across Canada’s boreal forest and to explore the drivers behind observed trends. Our results show that the majority of NDVI changes in Canada’s boreal forest reflect disturbance-recovery dynamics not climate change impacts, that greening and browning trends outside of disturbed forests are consistent with expected ecological responses to regional changes in climate, and that observed NDVI changes are geographically limited and relatively small in magnitude. By examining covariance between changes in NDVI and temperature and precipitation in locations not affected by disturbance, our results isolate and characterize the nature and magnitude of greening and browning directly associated with climate change. Consistent with biogeographic theory, greening and browning unrelated to disturbance tended to be located in ecotones near boundaries of the boreal forest bioclimatic envelope. We observed greening to be most prevalent in Eastern Canada, which is more humid, and browning to be most prevalent in Western Canada, where forests are more prone to moisture stress. We conclude that continued long-term climate change has the potential to significantly alter the character and function of Canada’s boreal forest, but recent changes have been modest and near-term impacts are likely to be focused in or near ecotones.

Nonlinear, Non-stationary and Seasonal Time Series Forecasting Using Different Methods Coupled with Data Preprocessing

Western South Africa is highly vulnerable to climate change due to its arid and semi-arid conditions, shaped by the region’s complex meteorology. This study investigated long-term patterns of rainfall, temperature variability, and drought dynamics in western South Africa from 1962 to 2022, with an emphasis on their implications for vegetation density. Meteorological data from the South African Weather Service and the Global Surface Summary of the Day were used to compute multi-timescale Standardized Precipitation and Evapotranspiration Index (SPEI), while Normalized Difference Vegetation Index (NDVI) time series (1992–2022) were derived from Landsat and MODIS satellite imagery processed on Google Earth Engine and ArcGIS. The findings reveal a significant and sustained decline in rainfall, particularly after 2000, accompanied by rising temperatures and an increase in drought frequency, duration, and severity. Notably, a single, prolonged drought from 2014 to 2022—lasting 7.5 years—marked the most severe event during the 60-year period. This drought had a region-wide impact on vegetation, with NDVI anomalies showing widespread vegetation browning in the western sub-region and relative vegetation greening in the eastern. This indicates that the impact was more pronounced in the western than in the eastern sub-region. While NDVI was positively correlated with rainfall (r = 0.37) and negatively with temperature (r = − 0.32), the modest strength of these relationships suggests that non-climatic factors may also influence vegetation cover. The study highlights the need for further investigation into the roles of land use, soil properties, and anthropogenic pressures in shaping vegetation responses to climate change.

Time series Normalized Difference Vegetation Index (NDVI) fitting, which describes the discrete time series observations with a mathematical model, can remove the residual noise, fill in the missing data, and produce a noise-free continuous time series; therefore, numerous time series fitting methods have been developed, but most of them focused on regularly sampled, noise-suppressed data, as NDVI time series derived from medium-spatial remote sensing images has irregular sampled and distortion-intensive characteristic, limiting the applicability of traditional methods. To address this problem, an automatic distortion-suppressed time series fitting method for irregular sampled NDVI is proposed in this paper. First, the observation quality is evaluated: (I) the observations on a fully clear image are identified as clear observations. (II) According to the prior knowledge that the NDVI is underestimated, the observations in the upper convex hull and above the line of two temporal-adjacent clear observations (identified by fully clear image and upper convex) are labelled as clear observations. (III) As the temporal evolution of NDVI is a slow and continuous process, deep-V observations, which has experienced an abrupt decrease and increase consecutively, are identified as noisy observations. Second, different weights are assigned to observations based on the quality assessment result, and the NDVI time series is weighted fitted. To highlight the observations above the fitting line and suppress those below the fitting line, the weights are updated according to the direction and distance between the actual observation and fitting prediction value. Then, fitting and reweighting processes are repeated until a stable fitting result is obtained. We use images acquired by the Multi-Spectral Imager (MSI) onboard Sentinel-2 satellite in Shouxian, Anhui Province to construct NDVI time series and test the method. The result demonstrates that our method can suppress the distortion and realize time series fitting automatically, compared with the state of art method, our method can obtain reasonably overestimation of the NDVI, providing an alternative for fitting irregular sampled NDVI.

In this study, the Savitzky-Golay filter was applied to smooth observed unnatural variations in the temporal profiles of the Normalized Difference Vegetation Index (NDVI) and the Enhanced Vegetation Index (EVI) time series from the MODerate Resolution Imaging Spectroradiometer (MODIS). We computed two sets of land cover classifications based on the NDVI and EVI time series before and after applying the Savitzky-Golay filter. The resulting classification from the filtered versions of the vegetation indices showed a substantial improvement in accuracy when compared to the classifications from the unfiltered versions. The classification by the EVIsg had the highest K (0.72) for all classes compared to those of the EVI (0.67), NDVI (0.63), and NDVIsg (0.62). Therefore, we conclude that the EVIsg is best suited for land cover classification compared to the other data sets in this study.

Accurate mapping of the spatio-temporal dynamic information of conversion from -ropland into forest/grassland is necessary for studying hydro-ecological effects of implementation of Grain to Green Project (GTGP) on the Loess Plateau, China. Currently, the accuracy of change detection of farmland and forest/grassland at 30-m scale in this area is seriously affected by insufficient temporal information from observations and irregular fluctuations in vegetation greenness caused by precipitation and human activities. In this study, an innovative method for continuous change detection of cropland and forest/grassland using all available Landsat time-series data. The period with vegetation coverage is firstly identified using normalized difference vegetation index (NDVI) time series. The intra-annual NDVI time series is then developed at a 1-day resolution based on linear interpolation and S-G filtering using all available NDVI data during the period when vegetation types are stable. Vegetation type change is initially detected by comparing the NDVI of intra-annual composites and the newly observed NDVI. Finally, the time of change and classification for vegetation types are determined using decision tree rules developed using a combination of inter-annual and intra-annual NDVI temporal metrics. Validation results showed that the change detection was accurate, with an overall accuracy of 88.9% ± 1.0%, and a kappa coefficient of 0.86, and the time of change was successfully retrieved, with 85.2% of the change pixels attributed to within a 2-year deviation.

Information on mangrove species composition and distribution is key to studying functions of mangrove ecosystems and securing sustainable mangrove conservation. Even though remote sensing technology is developing rapidly currently, mapping mangrove forests at the species level based on freely accessible images is still a great challenge. This study built a Sentinel-2 normalized difference vegetation index (NDVI) time series (from 2017-01-01 to 2018-12-31) to represent phenological trajectories of mangrove species and then demonstrated the feasibility of phenology-based mangrove species classification using the random forest algorithm in the Google Earth Engine platform. It was found that (i) in Zhangjiang estuary, the phenological trajectories (NDVI time series) of different mangrove species have great differences; (ii) the overall accuracy and Kappa confidence of the classification map is 84% and 0.84, respectively; and (iii) Months in late winter and early spring play critical roles in mangrove species mapping. This is the first study to use phonological signatures in discriminating mangrove species. The methodology presented can be used as a practical guideline for the mapping of mangrove or other vegetation species in other regions. However, future work should pay attention to various phenological trajectories of mangrove species in different locations.

Phenological dynamics of vegetation is considered as an important biological indicator for understanding the functioning of terrestrial ecosystems. Land surface phenology (LSP), the study of vegetation phenology from time series of vegetation indices (IV), has provided a comprehensive overview of ecosystem dynamics. Iberian Peninsula is one of the regions with the greatest diversity of ecosystems in European continent. It is therefore an excellent study area for monitoring phenological dynamics of vegetation. The aim of this study is to analyse the spatial variability of the phenology of the vegetation of the Iberian Peninsula and Balearic Islands for the period 2001-2017. NDVI (Normalized Difference Vegetation Index) time series were generated from the surface reflectance product MOD09Q1 at a spatial resolution of 250 meters and with a composite period of 8 days. Atmospheric disturbances and noise were reduced using a Savitzky-Golay smoothing filter. Different phenological metrics or phenometrics were extracted using a threshold-based method. Results showed the existence of a different behaviour between spring and autumn phenophases in the Atlantic and Mediterranean biogeographic regions. The Mediterranean mountainous areas showed a similar phenological behaviour to the Atlantic vegetation. Biogeographic regions showed an internal variability, which may be derived from the different behaviour of land covers (e.g., natural vegetation vs. crops).