Normalized Difference Vegetation Index Saturation Research Articles

The novel triangular spectral indices for characterizing winter wheat drought

Agricultural drought threatens food security and agricultural sustainable development. There have been numerous spectral indices from remote sensing images developed for monitoring crop drought. However, most present spectral indices are focusing on crop growth and Land Surface Temperature (LST), and the crop canopy water content are in less consideration simultaneously. Additionally, the Normalized Difference Vegetation Index (NDVI) is used for characterizing crop growth in almost all spectral drought indices, with the spectral saturation problem of NDVI for closed crop canopy. When vegetation cover is high, NDVI values tend to saturate, which makes them insensitive to further changes in crop health. Therefore, the NDVI saturation phenomenon may lead to an underestimation of the extent of crop drought, as it is not effective in identifying subtle changes in crops under high-density vegetation conditions. Hence, we propose three novel triangular spectral indices for characterizing winter wheat drought using three features including LAI, Land Surface Water Index (LSWI) and LST. For validating the proposed spectral indices, we compared the agreement between these indices with measured Relative Soil Moisture (RSM) and Volumetric Water Content (VWC) of soil in agricultural meteorological station and present popular indices including Crop Water Stress Index (CWSI), Temperature-Vegetation Drought Index (TVDI), and Vegetation Health Index (VHI). The results revealed that our proposed indices including Euclidean distance Crop Health Index (ECHI), Difference Crop Health Index (DCHI) and Perpendicular Water Stress Index (PWSI) outperformed the popular CWSI, TVDI and VHI, with stronger correlations with measured RSM and VWC in agricultural meteorological station. Secondly, there are spatial consistencies for characterizing winter wheat drought between proposed ECHI, DCHI and PWSI with popular CWSI, TVDI and VHI. In addition, our proposed ECHI, DCHI and PWSI have achieved good performance of drought monitoring both in irrigated and rainfed croplands. All these results suggest that our proposed indices have great potential in crop drought monitoring.

Integrating three-dimensional greenness into RSEI improved the scientificity of ecological environment quality assessment for forest

Normalized Difference Vegetation Index (NDVI) is widely used to represent the greenness indicator for the Ecological Environment Quality (EEQ) assessment based on the traditional Remote Sensing Ecological Index (RSEI). However, NDVI saturation issues are reported in agriculture and forest ecosystems at high greenness and biomass, creating a challenge when using NDVI to reflect the greenness components for forest EEQ assessment. In this paper, three-dimensional greenness (TDG) was obtained by the Forest Canopy Height (FCH) and Fractional Vegetation Cover (FVC) to quantify the forest greenness. The NDVI of RSEI was replaced with TDG to establish an improved remote sensing ecological index (TDRSEI) for the forest EEQ evaluation in the Central Yunnan, China. Moreover, we analyzed the difference between RSEI and TDRESI by qualitative and quantitative means and discussed deeply the optical saturation of NDVI by the quadratic function. The results shown that there were similar spatial distribution patterns and a strong correlation between TDG and FCH, Leaf Area Index (LAI), FVC, and NDVI, the TDG can be used to replace NDVI for reflecting forest greenness. The standard deviation of TDRSEI from the same FCH pixels was all less than that of RSEI, the absolute correlation coefficient between TDRSEI and ecological components were all greater than 0.63, and the mean values of TDRSEI (0.73, 0.84, and 0.90) from the relatively high FCH (20 m, 25 m, and 30 m) were greater than RSEI (0.71, 0.78, and 0.85), showing that the TDRSEI were more stable than RSEI for forest EEQ assessment and it can improve the EEQ in the high FCH. In Central Yunnan, China, the forest EEQ from TDRSEI and RSEI maps both increased with the growth of FCH, the distribution of RSEI had no significant difference between the low FCH and the high FCH, while the mean values of TDRSEI increased linearly with FCH growth. Besides, the ecological saturation points of TDRSEI and RSEI corresponding to the FCH were 19.50 m and 34.02 m, respectively. Therefore, TDRSEI method integrates the three-dimensional greenness to evaluate the forest EEQ objectively.

Using NDVI-NSSI feature space for simultaneous estimation of fractional cover of non-photosynthetic vegetation and photosynthetic vegetation

The NDVI-NSSI feature space, which is constructed by the NPV soil separation index (NSSI) calculated from the two near-infrared bands and the normalized difference vegetation index (NDVI), is often represented as a triangle. It can be combined with the pixel decomposition method to achieve simultaneous estimation of the fractional cover of non-photosynthetic vegetation, photosynthetic vegetation, and bare soil (fNPV, fPV, and fBS, respectively). However, NDVI saturation in dense vegetation and the deleterious effects of shading due to topography are inevitable. To overcome this difficulty, we compare and analyze the NDVI-NSSI and enhanced vegetation index-NSSI (EVI-NSSI) triangular spaces and their differences when applying on the Sentinel-2A/B images. The results show that EVI-NSSI index combination still forms a triangle-like feature, with the triangle transformed from obtuse to acute, and it can be better used for pixel three-decomposition when PV cover is high. NDVI-NSSI underestimates (overestimates) fPV when PV cover is low (high). The NDVI-NSSI and EVI-NSSI produce significantly different estimates of fPV and fBS but similar estimates of fNPV and can be combined for estimation of a wider range of NPV cover in different seasons, including sparse or dense green vegetation scenarios.

The Impact of Non-Photosynthetic Vegetation on LAI Estimation by NDVI in Mixed Grassland

Leaf area index (LAI) is widely used for algorithms and modelling in the field of ecology and land surface processes. At a global scale, normalized difference vegetation index (NDVI) products generated by different remote sensing satellites, have provided more than 40 years of time series data for LAI estimation. NDVI saturation issues are reported in agriculture and forest ecosystems at high LAI values, creating a challenge when using NDVI to estimate LAI. However, NDVI saturation is not reported on LAI estimation in grasslands. Previous research implies that non-photosynthetic vegetation (NPV) reduces the accuracy of LAI estimation from NDVI and other vegetation indices. A question arises: is the absence of NDVI saturation in grasslands a result of low LAI value, or is it caused by NPV? This study aims to explore whether there is an NDVI saturation issue in mixed grassland, and how NPV may influence LAI estimation by NDVI. In addition, in-situ measured plant area index (PAI) by sensors that detect light interception through the vegetation canopy (e.g., Li-cor LAI-2000), the most widely used field LAI collection method, might create bias in LAI estimation or validation using NDVI. Thus, this study also aims to quantify the contribution of green vegetation (GV) and NPV on in-situ measured PAI. The results indicate that NDVI saturation (using the portion of NDVI only contributed by GV) exists in grassland at high LAI (LAI threshold is much lower than that reported for other ecosystems in the literature), and that the presence of NPV can override the saturation effects of NDVI used to estimate green LAI. The results also show that GV and NPV in mixed grassland explain, respectively, the 60.33% and 39.67% variation of in-situ measured PAI by LAI-2000.

Smartphone near infrared monitoring of plant stress

The most widely used method for monitoring plant stress is the use of near infrared (NIR) spectrophotometry to calculate normalized difference vegetation index (NDVI), as defined by [NIR reflectance − red reflectance]/[NIR reflectance + red reflectance]. NDVI measures the chlorophyll absorption in the red spectrum relative to the scattering by cellular structure in NIR, and has been used to monitor vegetation health and subsequently its stress from aerial or satellite images. Rather than using an NIR spectrophotometer or an NIR camera that is rather expensive, we attempted to use a commercial smartphone, utilizing its (potentially unintended) ability in recognizing near infrared (NIR) color. Some of the most recent versions of smartphones have eliminated the NIR block filters on their cameras, and able to recognize NIR in their red pixels of CMOS array. Through attaching an inexpensive high pass filter at 800 nm to a smartphone camera, we were able to collect the NIR reflectance (with a high pass optical filter) and the red reflectance (without a filter), enabling NDVI assessments. This method was verified by measuring the NDVI values from a series of chlorophyll solutions, and showed a strong linear correlation with R2 = 0.948, corroborating the smartphone’s ability in evaluating NDVI. Using the leaves from three different plant species, the NDVI values were evaluated using the smartphone and compared with the plants’ chlorophyll contents using acetone extraction and subsequent spectrophotometry. A good linear relationship was found with R2 = 0.88–0.92. We further evaluated the NDVI values against the plants’ water contents (measured by oven-drying), showing the non-linear relationship with the NDVI saturation above 50% water content. The assay time was almost instantaneous, requiring only a smartphone and a high pass filter, thus allowing inexpensive, easy-to-use, rapid, and early prediction of plant stress that can be used for field and household applications.

Evaluating the utility of solar-induced chlorophyll fluorescence for drought monitoring by comparison with NDVI derived from wheat canopy

Normalized Difference Vegetation Index (NDVI) has been extensively used in continuous and long-term drought monitoring over large-scale, but with late response to drought-related changes of photosynthesis. Instead, solar-induced chlorophyll fluorescence (SIF) is more closely related to photosynthesis and thus is proposed to track the impacts of drought on vegetation growth. However, the detailed difference between SIF and NDVI in responding to drought has not been thoroughly explored. Here we present continuous ground measurements of NDVI and SIF at 760nm over four plots of wheat with different intensities of drought (well-watered treatment, moderate drought, severe drought and extreme drought). The average values of seasonal SIF were significantly lower under severe drought and extreme drought, while NDVI means only showed significant reduction in extreme drought. In the seasonal patterns, daily SIF could clearly separate the difference of drought gradient, while the difference of daily NDVI was clearer in the end of the field campaign. Daily SIF also significantly and positively correlated with soil moisture, indicating that SIF could be considered as an estimator of soil moisture to detect the information about agricultural drought. Furthermore, in extreme drought plot, the correlation of SIF and soil moisture was higher than that of NDVI and soil moisture in a shorter time lag (<15-day) but weaker in a longer time lag (longer than 30-day). The relationships of growth parameters with SIF and NDVI were further analyzed, showing a saturation of NDVI and unsaturation of SIF at high values of leaf area index and relative water content. These results suggested that SIF is better fit in early drought monitoring, especially over closure canopy, while NDVI is more feasible when drought lasted over a long time scale. Our findings in the study might provide deep insight into the utility of SIF in drought monitoring.

Dating lava flows of tropical volcanoes by means of spatial modeling of vegetation recovery

AbstractThe age of past lava flows is crucial information for evaluating the hazards and risks posed by effusive volcanoes, but traditional dating methods are expensive and time‐consuming. This study proposes an alternative statistical dating method based on remote sensing observations of tropical volcanoes by exploiting the relationship between lava flow age and vegetation cover. First, the factors controlling vegetation density on lava flows, represented by the normalized difference vegetation index (NDVI), were investigated. These factors were then integrated into pixel‐based multi‐variable regression models of lava flow age to derive lava flow age maps. The method was tested at a pixel scale on three tropical African volcanoes with considerable recent effusive activity: Nyamuragira (Democratic Republic of Congo), Mt Cameroon (Cameroon) and Karthala (the Comoros). Due to different climatic and topographic conditions, the parameters of the spatial modeling are volcano‐specific. Validation suggests that the obtained statistical models are robust and can thus be applied for estimating the age of unmodified undated lava flow surfaces for these volcanoes. When the models are applied to fully vegetated lava flows, the results should be interpreted with caution due to the saturation of NDVI. In order to improve the accuracy of the models, when available, spatial data on temperature and precipitation should be included to directly represent climatic variation. Copyright © 2017 John Wiley &amp; Sons, Ltd.

Evaluation of the Airborne CASI/TASI Ts-VI Space Method for Estimating Near-Surface Soil Moisture

High spatial resolution airborne data with little sub-pixel heterogeneity were used to evaluate the suitability of the temperature/vegetation (Ts/VI) space method developed from satellite observations, and were explored to improve the performance of the Ts/VI space method for estimating soil moisture (SM). An evaluation of the airborne ΔTs/Fr space (incorporated with air temperature) revealed that normalized difference vegetation index (NDVI) saturation and disturbed pixels were hindering the appropriate construction of the space. The non-disturbed ΔTs/Fr space, which was modified by adjusting the NDVI saturation and eliminating the disturbed pixels, was clearly correlated with the measured SM. The SM estimations of the non-disturbed ΔTs/Fr space using the evaporative fraction (EF) and temperature vegetation dryness index (TVDI) were validated by using the SM measured at a depth of 4 cm, which was determined according to the land surface types. The validation results show that the EF approach provides superior estimates with a lower RMSE (0.023 m3·m−3) value and a higher correlation coefficient (0.68) than the TVDI. The application of the airborne ΔTs/Fr space shows that the two modifications proposed in this study strengthen the link between the ΔTs/Fr space and SM, which is important for improving the precision of the remote sensing Ts/VI space method for monitoring SM.

Multi-Temporal Detection of Rice Phenological Stages Using Canopy Spectrum

Information on rice phenology is essential for yield estimation and crop management. To test the ability of remote sensing in detecting multiple phenological stages, paddy rice canopy spectrum was measured by a hand-held radiometer. Normalized difference vegetation index (NDVI) was calculated from spectrum, and the slope of NDVI was obtained as its difference. We evaluated the response of NDVI and its slope to rice growth with a comparison of two late-season rice cultivars. The results showed that NDVI and its slope curves had distinct variation corresponding to rice development and they could be used as cultivar-independent phenological indicators. The dates of flooding and transplanting, tillering, panicle development, heading and flowering, maturity, harvest stages, and even field management practices, could be deduced from these indicators. ‘NDVI ≤ 0’ could be used as a single threshold for the detection of flooding and transplanting. The largest spike in the curve of the NDVI slope indicated the duration of tillering stage. The next spike corresponded to panicle development. The heading and flowering stage was characterized by the maximum NDVI and the change of NDVI slope from positive to negative. At the maturity stage, NDVI decreased continuously, and its slope fluctuated just below zero. When rice grains were completely mature and ready for harvest, NDVI decline was accelerated. At harvest, NDVI slope reached its minimum value. The distinction between heading and maturity stages was obscure, most likely due to NDVI saturation at high biomass. The study might provide references for paddy rice phenology determination through remote sensing images.

NDVI saturation adjustment: A new approach for improving cropland performance estimates in the Greater Platte River Basin, USA

In this study, we developed a new approach that adjusted normalized difference vegetation index (NDVI) pixel values that were near saturation to better characterize the cropland performance (CP) in the Greater Platte River Basin (GPRB), USA. The relationship between NDVI and the ratio vegetation index (RVI) at high NDVI values was investigated, and an empirical equation for estimating saturation-adjusted NDVI (NDVIsat_adjust) based on RVI was developed. A 10-year (2000–2009) NDVIsat_adjust data set was developed using 250-m 7-day composite historical eMODIS (expedited Moderate Resolution Imaging Spectroradiometer) NDVI data. The growing season averaged NDVI (GSN), which is a proxy for ecosystem performance, was estimated and long-term NDVI non-saturation- and saturation-adjusted cropland performance (CPnon_sat_adjust, CPsat_adjust) maps were produced over the GPRB. The final CP maps were validated using National Agricultural Statistics Service (NASS) crop yield data. The relationship between CPsat_adjust and the NASS average corn yield data (r=0.78, 113 samples) is stronger than the relationship between CPnon_sat_adjust and the NASS average corn yield data (r=0.67, 113 samples), indicating that the new CPsat_adjust map reduces the NDVI saturation effects and is in good agreement with the corn yield ground observations. Results demonstrate that the NDVI saturation adjustment approach improves the quality of the original GSN map and better depicts the actual vegetation conditions of the GPRB cropland systems.

Estimating CO2 Sequestration by Forests in Oita Prefecture, Japan, by Combining LANDSAT ETM+ and ALOS Satellite Remote Sensing Data

CO2 sequestration of the forests in Oita Prefecture, Japan, was estimated using satellite remote sensing data. First, hybrid classification of the optical LANDSAT ETM+ data was performed using GIS to produce a detailed land cover map. CO2 sequestration for each forest type was calculated using the sequestration rates per unit area multiplied by the forest areas obtained from the land cover map This results in 3.57 MtCO2/yr for coniferous, 0.77 MtCO2/yr for deciduous broadleaf, and 2.25 MtCO2/yr for evergreen broadleaf, equivalent to a total of 6.60 MtCO2/yr for all the forest covers in Oita. Then, two different methodologies were used to improve these estimates by considering tree ages: the Normalized Difference Vegetation Index (NDVI) and the stem volume methods. Calculation using the NDVI method shows the limitation of this method in providing detailed estimations for trees older than 15 years, because of NDVI saturation beyond this age. In the stem volume method, tree ages were deduced from stem volume values obtained by using PALSAR backscattering data. Sequestration based on tree age forest subclasses yields 2.96 MtCO2/yr (coniferous) and 0.31 MtCO2/yr (deciduous broadleaf forests). These results show the importance of using not only detailed forest types, but also detailed tree age information for more realistic CO2 sequestration estimates. In so doing, overestimation of the sequestration capacity of forests could be avoided, and the information on the status and location of forest resources could be improved, thereby leading to sounder decision making in sustainable management of forest resources.

Development of regional wheat VI-LAI models using Resourcesat-1 AWiFS data

The time of forcing of spatial LAI to crop models at single or multiple stages is important to simulate crop biomass and yield in varying agro-climatic conditions and scales. The high temporal resolution (5-day) by Advanced Wide Field Sensor (AWiFS) on-board Resourcesat-1 Satellite IRS-P6 with 56 m spatial resolution and large swath (740 km) has substantially increased the availability of regional clear sky optical remote sensing data. The present study aimed at developing empirical vegetation index VI-LAI models for wheat using AWiFS optical data in four bands and in-situ measurements sampled over five different agro-climatic regions (ACRs) during 2005–2006 followed by validation during 2006–2007. While nonlinear relations exist for all the three normalized indices such as normalized difference vegetation index (NDVI), normalized difference water index (NDWI) and Green NDVI, linear relation was the best fit for ratio vegetation index (RVI). Both NDVI and RVI models generally showed better correlation ranges (0.65–0.84 for NDVI and 0.37–0.76 for RVI) than other indices. The common NDVI-LAI model was found to produce lower root mean square errors (RMSE) between 0.5 and 1.1 from pooled model than those between 0.5 and 1.32 from regional models. The rate of substantial increase in errors from NDVI-LAI model (RMSE of modeled LAI: 0.85 to 1.28) as compared to RVI-LAI model (RMSE of modeled LAI: 1.12 to 1.17) at LAI greater than 3, than below 3 revealed the early saturation of NDVI than RVI. It is therefore recommended that LAI estimates can be used to force crop simulation model upto early vegetative stage based on NDVI and maximum vegetative to reproductive stages based on RVI.

Comparison of Different Ground-Based NDVI Measurement Methodologies to Evaluate Crop Biophysical Properties

The usage of vegetation indices such as the Normalized Difference Vegetation Index (NDVI) calculated by means of remote sensing data is widely spread for describing vegetation status on large space scale. However, a big limitation of these indices is their inadequate time resolution for agricultural purposes. This limitation could be overcome by the ground-based vegetation indices that could provide an interesting tool for integrating satellite-based value. In this work, three techniques to calculate the ground-NDVI have been evaluated for sugar beet cultivated in South Italy in all its phenological phases: the NDVI1 based on hand made reflectance measurements, the NDVI2 calculated on automatically reflectance measurements and the broadband NDVIb based on Photosynthetically Active Radiation (PAR) and global radiation measurements. The best performance was obtained by the NDVIb. Moreover, crop-microclimate-NDVI relations were investigated. In particular, the relationship between NDVI and the Leaf Area Index (LAI) was found logarithmic with a saturation of NDVI at LAI around 1.5 m2 m-2. A clear relation was found between NDVI and crop coefficient Kc experimentally determined by the ratio between actual and reference measured or modelled evapotranspirations, while the relation between NDVI and crop actual evapotranspiration was very weak and not usable for practical purposes. Lastly, no relationship between the microclimate and the NDVI was found.

On the relationship of NDVI with leaf area index in a deciduous forest site

Numerous studies have reported on the relationship between the normalized difference vegetation index (NDVI) and leaf area index (LAI), but the seasonal and annual variability of this relationship has been less explored. This paper reports a study of the NDVI–LAI relationship through the years from 1996 to 2001 at a deciduous forest site. Six years of LAI patterns from the forest were estimated using a radiative transfer model with input of above and below canopy measurements of global radiation, while NDVI data sets were retrieved from composite NDVI time series of various remote sensing sources, namely NOAA Advanced Very High Resolution Radiometer (AVHRR; 1996, 1997, 1998 and 2000), SPOT VEGETATION (1998–2001), and Terra MODIS (2001). Composite NDVI was first used to remove the residual noise based on an adjusted Fourier transform and to obtain the NDVI time-series for each day during each year. The results suggest that the NDVI–LAI relationship can vary both seasonally and inter-annually in tune with the variations in phenological development of the trees and in response to temporal variations of environmental conditions. Strong linear relationships are obtained during the leaf production and leaf senescence periods for all years, but the relationship is poor during periods of maximum LAI, apparently due to the saturation of NDVI at high values of LAI. The NDVI–LAI relationship was found to be poor ( R 2 varied from 0.39 to 0.46 for different sources of NDVI) when all the data were pooled across the years, apparently due to different leaf area development patterns in the different years. The relationship is also affected by background NDVI, but this could be minimized by applying relative NDVI. Comparisons between AVHRR and VEGETATION NDVI revealed that these two had good linear relationships ( R 2=0.74 for 1998 and 0.63 for 2000). However, VEGETATION NDVI data series had some unreasonably high values during beginning and end of each year period, which must be discarded before adjusted Fourier transform processing. MODIS NDVI had values greater than 0.62 through the entire year in 2001, however, MODIS NDVI still showed an “M-shaped” pattern as observed for VEGETATION NDVI in 2001. MODIS enhanced vegetation index (EVI) was the only index that exhibited a poor linear relationship with LAI during the leaf senescence period in year 2001. The results suggest that a relationship established between the LAI and NDVI in a particular year may not be applicable in other years, so attention must be paid to the temporal scale when applying NDVI–LAI relationships.