On the relationships between satellite-based drought index and gross primary production in the North Korean croplands, 2000–2012

Drought is one of the main constraints on vegetation growth and crop yields, although land ecosystems differ in their sensitivity to drought. Satellite-based vegetation indices, such as the Normalized Difference Vegetation Index (NDVI), have been used in many drought studies, but they may not sufficiently represent the water content of vegetated land. Alternatively, the Normalized Difference Drought Index (NDDI) has been developed by integrating the NDVI and Normalized Difference Water Index (NDWI). In this letter, we examine how drought affects crop growth by quantifying the relationships between NDDI and the Gross Primary Production (GPP) derived from the moderate-resolution imaging spectroradiometer. In the North Korean croplands, NDDI had a strong negative correlation with GPP during 2000–2012. The relationships were more significant under relatively dry conditions (e.g., dry seasons or dry regions). The impacts of NDDI on GPP was greater in summer than in spring, which indirectly shows summer drought may be more critical to crop productivity. The NDDI–GPP relationship was slightly time-lagged in spring, which indicates that vegetation productivity may not always respond instantly to surface dryness. The NDDI can be a viable option for measuring the impacts of drought on vegetation and agriculture over a wide area.     

Independent component analysis of MODIS-NDVI data in a large South American wetland

Monthly images of Normalized Difference Vegetation Index (NDVI) from the moderate resolution imaging spectroradiometer (MODIS) are used to characterize the spatio-temporal variability of vegetation in a large South American wetland (SAW) (located in the Paraná River floodplain) during the period 2000–2009. While these data do not meet the requirements of classical component extraction techniques (CETs) (e.g. principal component analysis (PCA)), they are suitable for the modern method named independent component analysis (ICA). Hence, ICA is used here to extract three statistically independent modes of inter-annual MODIS-NDVI variability that are successfully interpreted as vegetation responses to hydrological changes. One mode isolates the vegetation response to a severe drought associated with La Niña 2007–2008. Another component reflects the expansion (or contraction) of lagoons owing to high (or low) water level of the Paraná River. The remaining mode captures the vegetation decrease caused by the flood related to El Niño 2006–2007. The results presented here for a particular wetland suggest that ICA of NDVI images is a powerful tool for identifying the physical causes of vegetation changes in other large wetlands.     

Sentinel-2 based prediction of spruce budworm defoliation using red-edge spectral vegetation indices

ABSTRACT

This research compares the capabilities of various Sentinel-2-derived spectral vegetation indices (SVIs) in particular red-edge SVIs to detect and classify spruce budworm (Choristoneura fumiferana) (SBW) defoliation using Support Vector Machine (SVM) and Random Forest (RF) models. The results showed the superiority of RF in model building for defoliation detection and classification into three classes (nil, light, and moderate) with overall errors of 17% and 32%, respectively. The most important variables for the best model were Enhanced Vegetation Index 7 (EVI7), Modified Chlorophyll Absorption in Reflectance Index (MCARI), Inverted Red-Edge Chlorophyll Index (IRECI), Normalized Difference Infrared Index 11 (NDII11) and Modified Simple Ratio (MSR). Red-edge SVIs were more effective variables for light defoliation detection compared to traditional SVIs such as Normalized Difference Vegetation Index (NDVI) and EVI8. These findings can help improve current remote sensing-based SBW defoliation detection and monitoring.     

Optimizing harmonics from Landsat time series data: the case of mapping rainfed and irrigated agriculture in Zimbabwe

ABSTRACT

Explicit and large-scale information on farming systems is important for many applications such as crop production estimates, drought impact assessments or water footprint analysis. This contribution investigated the possibility of optimizing harmonic functions fitted to cloud-corrected Landsat NDVI (Normalized Difference Vegetation Index) time-series imagery to effectively map rainfed and irrigated croplands in Zimbabwe. The following harmonic optimizations were investigated: the effect of the linear trend, length of the time series data and the harmonics degree. The verification accuracy scores for mapping the two farming systems were used to ascertain the most optimal harmonics settings. The most accurate classification results (overall accuracy of 97%) were produced when the linear trend was excluded and the full time period (2013 to 2018) with a 7^th degree harmonics fitting function was used. Optimized harmonics provide an effective way to compute vegetation seasonality signals while accounting for data gaps and residual noise inherent in Landsat time-series data.     

Assessing seasonal trends and variability of vegetation growth from NDVI3g,MODIS NDVI and EVI over South Asia

The characterization of vegetation dynamics over South Asia (SA) has been primarily conducted using satellite time series of Advanced Very High Resolution Radiometer (AVHRR) Normalized Difference Vegetation Index (NDVI). However, various vegetation indices may show diverse trend patterns over the same area. This study analysed the consistency of the vegetation spatiotemporal trends from AVHRR version 3 NDVI (NDVI3g) with the Moderate Resolution Imaging Spectroradiometer (MODIS) NDVI and Enhanced Vegetation Index (EVI) over SA during various seasons, assuming that MODIS products are of higher quality. Results showed that the spatiotemporal vegetation trends derived from the NDVI3g were analogous to both MODIS NDVI and EVI indicating greening over semi-arid regions where croplands dominate and browning over tropical/subtropical forest areas. Correlations among them were better during winter monsoon. Discrepancies occurred in tropical/subtropical densely vegetated (humid) and complex topographic areas specifically during summer monsoon (SM). This study improved the understanding of the heterogeneous vegetation trends over the vast complicated terrain of SA. It was revealed that NDVI3g is reliable to quantify vegetation trends over SA, however, calibration errors still could introduce biases during SM season.     

Increased grass NDVI under contrasting trends of precipitation change over North China during 1982–2011

Changes in precipitation patterns were expected to have strong impacts on temperate ecosystem dynamics. North China has experienced opposite trends of precipitation change (increased in the west and decreased in the east) in the past several decades. Under such a background, we analysed mean growing season (GS) (April–October) grass Normalized Difference Vegetation Index (NDVI) changes using combined dataset of Global Inventory, Monitoring, and Modelling Studies and Moderate Resolution Imaging Spectroradiometer NDVI in North China during 1982–2011. The results showed that in mean GS NDVI increased for grasslands in both Northeastern China (NE) and Northwestern China (NW). Increase in NDVI in NW was mainly due to the increase in precipitation (r = 0.50, p < 0.01). However, the decrease in precipitation did not cause a decrease in grass NDVI in NE, suggesting that precipitation is still higher than the most sensitive value and NDVI changes were significantly correlated with the increased temperature (r = 0.43, p < 0.05).     

Extraction of coastal raft cultivation area with heterogeneous water background by thresholding object-based visually salient NDVI from high spatial resolution imagery

The development of high spatial resolution satellite imaging has enabled the acquisition of mariculture area information. This data could play an important role in mariculture investigations, ocean disaster evaluations, and coastal management. Because chlorophyll is concentrated in the widely distributed raft culture (a major kind of mariculture), the Normalized Difference Vegetation Index (NDVI) can be used for extraction. However, extensive coastal raft culture is easily confused with the heterogeneous water background. This results in unsatisfactory extraction when surveying a large water area with heterogeneous water background. By combining object-based image analysis and the centre-surround mechanism of a visual attention model, we propose an object-based visually salient NDVI (OBVS-NDVI) feature. Comparison experiments using Gaofen-2 spectral imagery of Luoyuan Bay, Fuzhou, China, indicate that OBVS-NDVI can effectively discriminate raft cultivation areas over large areas with a heterogeneous water background.     

Seasonal variations in litter production and its relation with MODIS vegetation indices in a semi-deciduous forest of Mato Grosso

Litter production is related to canopy processes including the timing and amount of leaf development, reproduction and net primary production. However, quantifying spatial and temporal patterns in litter production is complicated in Amazonian semi-deciduous forests because of high spatial heterogeneity and seasonal variation in rainfall. Here, we use monthly measurements of litter production and Moderate Resolution Imaging Spectroradiometer (MODIS) Normalized Difference Vegetation Index (NDVI) and Enhanced Vegetation Index (EVI) composites over a 6-year period to assess whether MODIS can be used to quantify litter production of tropical semi-deciduous forests. Original MODIS NDVI and EVI values were poorly related to the seasonal periodicity in litter production, but after using singular spectrum analysis (SSA) signal extraction techniques, clear relationships between litter production and the NDVI and EVI emerged. These results indicate that MODIS NDVI and EVI data are useful for detecting temporal patterns in litter production for Amazonian semi-deciduous forests if signal extraction analyses such as SSA are conducted.     

Integrating Mahalanobis typicalities with a neural network for rubber distribution mapping

Accurate rubber distribution mapping is critical to the study of its expansion and to provide a better understanding of the consequences of land-cover and land-use change on carbon and water cycles. Employing Mahalanobis typicalities as inputs to a hard classifier to enhance the capability of generalization has not previously been explored. This letter presents a novel approach by integrating Mahalanobis typicalities with the multi-layer perceptron (MLP) neural network for mapping of rubber. A case study from the Thai–Lao and Sino–Lao borders was conducted using Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) data. Different combinations of the nine ASTER bands including Visible and Near Infrared (VNIR) and Short-wave Infrared (SWIR), Normalized Difference Vegetation Index (NDVI) and Mahalanobis typicalities were used as input variables to the MLP. Results indicate that including Mahalanobis typicalities as input variables can improve the MLP's performance and increase the user's accuracy of rubber mapping.     

A new narrow band vegetation index for characterizing the degree of vegetation stress due to copper: the copper stress vegetation index (CSVI)

This study proposed a new narrow band index to characterize the Cu (Copper) stress degree on vegetation (Copper Stress Vegetation Index, CSVI). Firstly, the spectral reflectance and biochemical data of wheat, pea, locust and ash were analysed using Pearson correlation coefficient (r) to select wavelengths sensitive to Cu stress. The calculated Pearson correlation coefficients suggested that the reflectance near 550 nm and 700 nm correlated positively with Cu contents in leaves and solutions, and negative correlation was present in the range of 800–900 nm. Secondly, the selected wavelengths of 550 nm, 700 nm, and 850 nm were used to establish CSVI, and it was compared with existing popular vegetation indices (VIs) related to heavy metal stress (Normalized Difference Vegetation Index (NDVI), Red-Edge Position (REP), Difference Vegetation Index (DVI), Photochemical Reflectance Index (PRI)) by calculating Pearson correlation coefficient between VIs and Cu contents in leaves and solutions. Thirdly, verifications of CSVI on other vegetations were conducted, and the performance of CSVI was also compared with that of NDVI, REP, DVI, and PRI. The results suggested that CSVI showed significant correlation with Cu stress degree, and the correlation of CSVI was much stronger than that of other VIs for all the tested vegetations. The proposed CSVI characterizes the Cu stress degree on vegetation with advantages of better effectiveness, straightforward calculation, and robustness for different vegetations. This study focused on the spectral reflectance at the leaf scale, so it is expected that future work extends it to canopy scale and mixed-pixel scale.     

Multi-dimensional histogram-based information capacity analysis of urban heat island effect using Landsat 8 data

The vegetation and impervious surface area (ISA) are the most key indicators for the urban heat islands (UHI). This study used the normalized difference vegetation index (NDVI) as the indicator of vegetation abundance, the modified normalized difference impervious surface index (MNDISI) as the indicator of impervious surface fraction to estimate the land surface temperature (LST)–vegetation relationship and LST–ISA relationship. The land surface cover types were obtained by classification and regression tree (CART). Results demonstrate and verify that LST possessed strong negative correlations with NDVI and positive correlations with MNDISI at various spatial resolutions (30–960 m). Both correlation coefficients reached their strongest points at 30 m resolution, which is believed to be the operational scale of LST, NDVI and MNDISI. Further, information capacity (IC), as a spatial index, is used to characterize the spatial pattern of UHI. Results show that the IC of LST (LST_IC) possessed strong positive correlations with the IC of NDVI (NDVI_IC) and the IC of MNDISI (MNDISI_IC). It is suggested that the spatial pattern of UHI has a direct correspondence with the vegetation and ISA.     

A new estimate of tundra-biome phytomass from trans-Arctic field data and AVHRR NDVI

It is often assumed that the Normalized Difference Vegetation Index (NDVI) can be equated to aboveground plant biomass, but such a relationship has never been quantified at a global biome scale. We sampled aboveground plant biomass (phytomass) at representative zonal sites along two trans-Arctic transects, one in North America and one in Eurasia, and compared these data to satellite-derived NDVI. The results showed a remarkably strong correlation between total aboveground phytomass sampled at the peak of summer and the maximum annual NDVI (R ²?=?0.94, p < 0.001). The relationship was almost identical for the North America and Eurasia transects. The NDVI–phytomass relationship was used to make an aboveground phytomass map of the tundra biome. The approach uses a new and more accurate NDVI data set for the Arctic (GIMMS3g) and a sampling protocol that employs consistent methods for site selection, clip harvest and sorting and weighing of plant material. Extrapolation of the results to zonal landscape-level phytomass estimates provides valuable data for monitoring and modelling tundra vegetation.     

A new vegetation heavy metal pollution index for detecting the pollution degree of different varieties of maize under copper stress

This study proposed a new vegetation heavy metal pollution index VHMPI to detect the pollution degree of different varieties of maize under copper stress, which provides a new idea for the detection of heavy metal pollution in vegetation. In order to ensure the outdoor growth environment of maize, we put all maize into outdoor greenhouse. The spectral reflectance interval of 450 nm–850 nm of maize leaves was processed by the first order differential (D) and continuum removal (CR), and the DCR spectral curve was obtained. The Pearson correlation coefficient (R) was used to analyze the DCR data and the biochemical data and select characteristic bands that sensitive to heavy metal Cu. The calculated Pearson correlation coefficients suggested that the DCR value at 490 nm–520 nm and 680 nm–700 nm presented a linear positive correlation close to 1 with the Cu²⁺ contents in soil and leaves, and a linear negative correlation close to ?1 was present in the range of 630 nm–650 nm and 710 nm–750 nm. We selected the DCR value of wavelengths 505 nm, 640 nm, 690 nm and 730 nm to establish VHMPI, and compared it with conventional vegetation indices (VIs) by calculating Pearson correlation coefficient between them and Cu contents in soil and leaves, Vegetation indices include WBI (Water Band Index), PSNDa (Pigment Specific Normalized Difference a), PRI (Photochemical Reflectance Index), NDVI (Normalized Difference Vegetation Index). Maize leaf spectral data obtained from experiments in 2017 were used for verification, VHMPI was also compared with WBI, PSNDa, PRI and NDVI.The results suggested that VHMPI showed a significant correlation with Cu²⁺ stress concentration,and the correlation of VHMPI was much stronger than that of other vegetation indices. The proposed VHMPI detects the pollution degree of maize with different varieties and in different periods under copper stress has advantages of straightforward calculation, robustness, and high effectiveness. This study focused on the laboratory leaf scale, so it is expected that future work extends it to a wide range of field scale and image scale.     

Use of spatial autocorrelation and time series Landsat images for long-term monitoring of surface water shrinkage and expansion in Guanting Reservoir,China

ABSTRACT

Reservoirs are closely related to anthropic activities, and quantifying the long-term dynamics of surface water in reservoirs could be useful for decision-makers to improve the actual strategies of reservoir management. This study used the global Moran’s I index, modi?ed Normalized Difference Water Index (MNDWI) and a total of 596 Landsat images during 1985–2018 for tracking the annual dynamics of water extent in the process of water shrinkage and expansion in Guanting Reservoir, China. Landscape metrics related to the area, elongation, fragmentation, and edge complexity of surface water in reservoir landscape were computed for tracking the annual dynamics of surface water patterns. Statistical comparison between the results of global Moran’s I index and landscape metrics indicates that except for the complexity of water and non-water edge, global Moran’s I index can successfully estimate the dynamics of the area, elongation and fragmentation of surface water in the reservoir. This study proposed a continuous approach of long-term monitoring of surface water patterns using spatial autocorrelation that might be used in the areas where the surface water extraction is difficult and water dynamics are complex.     

MODIS BRDF effects over Brazilian tropical forests and savannahs: a comparative analysis

Bidirectional reflectance distribution function (BRDF) effects over Brazilian tropical forests (Amazon) and savannahs (Cerrado) were inspected for differences in high-quality pixel retrievals, view direction, view zenith angle (VZA), solar zenith angle (SZA) and relative azimuth angle (RAA). By comparing Moderate Resolution Imaging Spectroradiometer (MODIS) data, corrected and non-corrected for BRDF effects (2000–2014), we evaluated the magnitude of such effects over reflectance and vegetation indices (VIs). The VIs were the Enhanced Vegetation Index (EVI) and the Normalized Difference Vegetation Index (NDVI). From the Amazon to the Cerrado, we observed a higher frequency of high-quality pixel retrievals, a decrease in VZA, and an increase in seasonal SZA amplitude. Brightness increased in the backscattering direction and with shifts in RAA over tropical forests toward the BRDF hotspot at the end of dry season (September). Compared to the savannahs from the Cerrado or those from northern Amazon, stronger BRDF effects were observed for Amazonian tropical forests and for the EVI. BRDF changes in the dry season NDVI amplitude (NDVI_Sept minus NDVI_June) were lower than 5% for both biomes. For the EVI amplitude, we observed changes up to 20% over the savannahs and close to 60% for most of the Amazon. However, even after BRDF correction and in spite of the observed differences close to 0.02, EVI values from June and September were still statistically different from each other.     

Discrimination of areas infected with coffee leaf rust using a vegetation index

Coffee leaf rust is for the coffee industry potentially one of the causes of a sustainability crisis. Currently, on-site disease detection is the only effective method to fell coffee trees for prevention of the infection. However, accurate infection detection over wide areas is difficult when conducted by ground surveys. Here, we examine the application of a remote sensing method. The Normalized Difference Vegetation Index (NDVI) values of coffee farms were computed using satellite images and compared with the results of the ground truth. We found that the standard deviation of the NDVI value (σ_NDVI) in damaged farms increases as the average NDVI value decreases. This fact implies that the disease progresses in-homogeneously inside a damaged area. In the present analysis, up to 94.1% of the damaged farms were discriminated by combining the NDVI and σ_NDVI thresholds when 75.0% of the damaged farms had NDVI values under 0.732 and σ_NDVI over 0.044. Our monitoring method enabled us to take early-stage countermeasures against the infection, and it could be applied to other vegetation diseases.     

Combining ad hoc spectral indices based on LANDSAT-8 OLI/TIRS sensor data for the detection of plastic cover vineyard

In this article, we are proposing a method using Landsat-8 Operational Land Imager and Thermal Infrared Sensor data for agricultural plastic cover detection. Four normalized difference indices were combined in the procedure described to achieve consistent results: the green Normalized Difference Vegetation Index and three ad hoc spectral indices purposely created for this study (rescaled brightness temperature, Plastic Surface Index and Normalized Difference Sandy Index). The sampling time related to the preliminary collection of spectral information on plastic surfaces was reduced using information gathered through the Quality Assessment and Cloud Quality bands. The overall accuracies observed were on average higher than 80%,and the low cost of the open data set used, lacking ancillary data, demonstrated the reliability of the proposed method, proving its suitability for environmental and agricultural monitoring over large areas.     

Dieback classification modelling using high-resolution digital multispectral imagery and in situ assessments of crown condition

Quantifying dieback in forests is useful for land managers and decision makers seeking to explain spatial disturbances and understand the cyclic nature of forest health. Crown condition is assessed as reference to dieback in terms of the density, transparency, extent and in-crown distribution of foliage. At 20 sites in the Yalgorup National Park, Western Australia, a total of 80 Eucalyptus gomphocephala crowns were assessed both in situ (2008) and using two acquisitions (2008 and 2010) of airborne imagery. Each tree was assessed using four crown-condition indices: Crown Density, Foliage Transparency, the Crown Dieback Ratio and Epicormic Index combined into a single index called the Total Crown Health Index (TCHI). The airborne imagery is like value calibrated then classified and modelled using in situ canopy condition assessments resulting in a quantification of crown-condition change over time. Comparison of Normalized Difference Vegetation Index (NDVI), Soil-Adjusted Vegetation Index (SAVI) and a novel Red-Edge Extrema Index (REEI) suggests that the latter is more suited to classification applications of this type.     

Assessing vegetation cover on the date of satellite-derived start of spring

Time series of satellite imagery are commonly used to study and model phenology. To use these models, their results must be compared with time series of areal field data, and vegetation condition must be assessed relative to model predictions. Field data and Moderate Resolution Imaging Spectroradiometer (MODIS) data for corn fields in Illinois, USA, were collected throughout the growing season, including vegetation cover fraction (VCF) derived from kite aerial photography (KAP). The mean height of corn on the estimated start of spring (SOS) date was just over 2 cm and the mean VCF on SOS was nearly 10%, indicating that satellite models of phenology lag behind field-based measures of phenology like crop emergence. The relationships between MODIS Normalized Difference Vegetation Index (NDVI) and both KAP NDVI (coefficient of determination (R ²)?=?0.918, p < 0.000) and KAP VCF (R ²?=?0.920, p < 0.000) were strong, highlighting the importance of areal field data in phenology studies.     

The potential of the MERIS Terrestrial Chlorophyll Index for crop yield prediction

Regional and global prediction of crop yield by remote sensing is of vital importance for food security in China. Vegetation indices (VIs) sensitive to leaf area index (LAI), such as Normalized Difference Vegetation Index (NDVI), have been widely used for crop yield prediction. However, chlorophyll content is the key component for crops to convert light energy to organics in the process of photosynthesis, and crop yield may be more accurately predicted by remote sensing models based on a spectral index sensitive to chlorophyll content, such as the medium resolution imaging spectrometer (MERIS) Terrestrial Chlorophyll Index (MTCI). In this study, we investigated the potential of MTCI for crop yield prediction. Firstly, the MTCI and the NDVI products in Henan Province, China, from 2003 to 2011 with a temporal resolution of half a month were calculated from the daily level-2 reduced resolution ENVISAT MERIS reflectance product (MER_RR_2P) using a Maximum Value Composite algorithm. Secondly, we established winter wheat prediction models based on MTCI and NDVI. Then, the accuracy of MTCI for winter wheat yield prediction was examined and compared to the NDVI through a leave-one-out cross-validation approach. The results show that (1) the correlation coefficient between yield and MTCI is significantly higher than that of models based on NDVI and the errors of models based on MTCI are lower than those of models based on NDVI except for milking stages. Moreover, the crop yield prediction model based on accumulated MTCI through the reviving stage to milking is most significantly correlated to crop yield with a coefficient of determination of 0.849 for MTCI; (2) the optimum phase for crop yield prediction based on MTCI is the heading stage, about 30 days earlier than that of NDVI (milking stage), which is another advantage of MTCI in crop yield prediction; and (3) the validation error of the crop yield model based on the accumulated MTCI is half of that based on the accumulated NDVI. The study indicates that MTCI is potentially a better VI for crop yield prediction compared to NDVI.