Spectroscopic detection of nitrogen concentrations in sagebrush

The ability to estimate foliar nitrogen in semi-arid landscapes can yield information on nutritional status and improve our limited understanding of controls on canopy photosynthesis. We examined two spectroscopic methods for estimating sagebrush dried leaf and live shrub nitrogen content: first derivative reflectance (FDR) and continuum removal. Both methods used partial least squares (PLS) regression to select wavebands most significantly correlated with nitrogen concentrations in the samples. Sagebrush dried leaf spectra produced PLS models (R ² = 0.76–0.86) that could predict nitrogen concentrations within the data set more accurately than PLS models generated from live shrub spectra (R ² = 0.41–0.63). Inclusion of wavelengths associated with leaf water in the FDR transformations appeared to improve regression results. These findings are encouraging and warrant further exploration into sagebrush reflectance spectra to characterize nitrogen concentrations.     

Estimation of particulate zinc using MERIS data of the Pearl River Estuary

Marine heavy metals in the particulate phase are considered to be undetectable from remote sensing since their minor spectral signal is masked by marine optical constitutes. However, over 80% of heavy metals adsorb to and complex with the suspended sediment and the quantity of adsorbed heavy metals strongly depends on suspended sediment load and size. Thus, the suspended sediment load and size could be useful information to estimate heavy metal levels through remote-sensing techniques. We investigate the potential to derive heavy metal concentration, using zinc as an example. Eight cruises in the Pearl River Estuary were carried out to measure zinc concentrations in the particulate phase (C _zn) and remote-sensing reflectance (R _rs) data from 2009 to 2010. The relationship of R _rs ratio in the near infrared bands and the scattering spectral slope (which indicates particle size distribution (PSD)) is first proposed. According to a previously established suspended sediment algorithm for the study area and the newly proposed relationship of remote-sensing reflectance and suspended sediment size, a piecewise algorithm is developed to retrieve C _zn. The accuracies of the algorithm developed here (correlation coefficient (r) = 0.88, mean relative error (MRE) = 24.7% and root mean square error (RMSE) = 5.51) are higher than those algorithms only considering suspended sediment load (r = 0.79, MRE = 35.3% and RMSE = 8.93) or size (r = 0.78, MRE = 30.0% and RMSE = 7.19). The algorithm is applied to Medium Resolution Imaging Spectrometer (MERIS) full resolution data to obtain C _zn distribution in the Pearl River Estuary. The C _zn from image is mainly in the range from 10 to 63 μg l^?1, which is consistent with the in situ data ranges (5.67–86.62 μg l^?1).     

Relationship between canopy height and Landsat ETM+ response in lowland Amazonian rainforest

This letter investigates the influence of within-pixel variation of canopy height on the spectral response recorded in Landsat Enhanced Thematic Mapper (ETM+) data for tropical rainforest. Forest canopy height is derived from airborne, small-footprint LiDAR data acquired using a Leica ALS50 II system. The field site is in the Tambopata National Reserve, in Peruvian Amazonia, where forest types include regenerating, swamp, floodplain and terra firme. For individual Landsat ETM+?bands, the strongest correlation for maximum, mean and standard deviation of canopy height occurred with ETM+?Band 4 (near infrared) for regenerating, floodplain and terra firme forest, and with ETM+?Band 5 (middle infrared) for swamp forest. For normalized difference band indices, ND42 and ND43 (i.e. the Normalized Difference Vegetation Index, NDVI) showed strong correlation with both mean and maximum canopy height for regenerating and terra firme forest, and with maximum and standard deviation of canopy height for floodplain forest. The palm-dominated swamp forest showed weaker relationships, with the strongest occurring for ND45 and ND52 with mean canopy height. Many papers have identified middle-infrared bands as being most sensitive to tropical rainforest structure, although these have often focussed on young regenerative forests. By focussing on older regenerative forest (of >25 years since land abandonment) and mature rainforest types, this work has shown that there is considerable variation with how structure may influence spectral reflectance and lends support to the hypothesis that canopy height distribution and shadowing effects caused by canopy complexity and the presence of emergent trees is what most significantly influences spectral response for tropical rainforests.     

Retrieving fPAR of maize canopy using artificial neural networks with airborne LiDAR and hyperspectral data

ABSTRACT

Accurate estimation of the fraction of absorbed photosynthetically active radiation (fPAR) is important for maize growth and yield estimations. Light detection and ranging (LiDAR)-derived canopy vertical structural and hyperspectral image-derived vegetation spectral information are complementary for vegetation fPAR estimation. This study explores the potential of artificial neural networks (ANNs) with two types of data to estimate maize fPAR. First, 45 metrics were derived from LiDAR data and 13 from a hyperspectral image. Then, the ANNs and stepwise multiple linear regression (SMLR) methods were used to estimate the fPAR. Finally, model validity was assessed using in-situ data. Results showed that the ANNs performed better in fPAR inversion (R ² = 0.910, adj. R ² = 0.921, RMSE = 0.046, RRMSE = 0.056, where R ² is the coefficient of determination, adj. R ² the adjusted coefficient of determination, RMSE the root mean squared error, and RRMSE the relative root mean squared error) than SMLR (R ² = 0.638, adj. R ² = 0.609, RMSE = 0.077, RRMSE = 0.092) and SMLR with the natural logarithm of data (R ² = 0.855, adj. R ² = 0.825, RMSE = 0.067, RRMSE = 0.081). This study is helpful for guiding the accurate estimation of maize fPAR using remote sensing techniques.     

The potential of Sentinel-2 data for estimating biophysical variables in a boreal forest: a simulation study

In this study we use ground reference data from 962 forest plots to demonstrate the potential of Sentinel-2 (S2) bands in estimating canopy biophysical properties in boreal forests in Finland. We simulated canopy bidirectional reflectance factors (BRFs) using the PARAS model, which applies photon recollision probability. Results showed that the highest correlation between simulated S2 BRFs and fraction of absorbed photosynthetically active radiation (fPAR) was for the band combination band 7/band 9 (wavelengths 773–793 nm and 935–955 nm, respectively) (the coefficient of determination (R²) was 0.93). For effective leaf area index (LAI_e) the best band combination was band 8/band 4 (wavelengths 785–900 nm and 650–680 nm, respectively) (R² = 0.93). Based on this study, the above-ground biomass (AGB) and S2 band combinations did not show strong relationships (R² = 0.24). The new inverted red-edge chlorophyll index (IRECI) and Sentinel-2 red-edge position – index (S2REP) showed moderate relationships with fPAR (R² = 0.61 and R² = 0.45, respectively) and LAI_e (R² = 0.56 and R² = 0.30, respectively). This study demonstrated the potential of the S2 data to estimate canopy biophysical properties.     

Estimating the suspended sediment concentration from TM/Landsat-5 images for the Araguaia River – Brazil

ABSTRACT

In this study, 68 images from TM/Landsat-5 sensor were used to estimate Suspended Sediment Concentration (SSC) along of the Araguaia River, Brazil. These were combined with in-situ SSC, hydrosedimentometric station (categorical variable), and remote sensing reflectance. Top-of-Atmosphere (ToA) and surface reflectance data were evaluated. Multiple regression models with ToA reflectance using VNIR bands, band ratios, SWIR band 5 as input and station as categorical variable were more accurate with adjusted coefficient of determination (adjusted R²) = 0.87 and normalized root mean square error (NRMSE) = 10.09% compared to the models with surface reflectance with adjusted R² = 0.60 and NRMSE = 15.43%. Results confirm the potential for estimation of SSC from TM/Landsat-5 historical series data between 1984 and 2012, for which in-situ database is rare. Based on this empirical model, future studies may provide better analysis of spatiotemporal variations of sediment transport along the Araguaia River with the SSC temporal series reconstitution.     

Estimating chlorophyll-a concentration based on a four-band model using field spectral measurements and HJ-1A hyperspectral data of Qiandao Lake,China

Accurate estimation of phytoplankton chlorophyll-a (chl-a) concentration from remote sensing data is challenging due to the complex optical properties of case II waters. Recently, a novel semi-analytical four-band model was developed to estimate chl-a concentration in turbid productive waters. The objective of this study was to evaluate the performance of the four-band model and extend its application to hyperspectral satellite data for estimating chl-a concentration in Qiandao Lake of China. Based on field spectral measurements and in situ water sampling, the four-band model expressed as [R_rs^?1(661.6) – R_rs^?1(706.7)] [R_rs^?1(714.8) – R_rs^?1(682.2)]^?1 was calibrated after band tuning, where R_rs^?1 represents the reciprocal of the remote sensing reflectance. The spectral-based four-band model accounted for more than 88% of variance in chl-a concentration with a root mean square error (RMSE) of 1.47 μg l^?1. To justify the potential of this model with satellite data, comparable wavelengths selected from HJ-1A Hyperspectral Imager (HSI) imagery were utilized to calibrate the four-band model. The HSI-based model explained about 80% of chl-a variation with an RMSE of 1.35 μg l^?1. Experimental results also showed that the four-band model outperformed its three-band counterpart. The results validated the rationale of the four-band model and demonstrated the effectiveness of this model for estimating chl-a concentration from both in situ spectral data and HJ-1A hyperspectral satellite imagery.     

Accounting for non-photosynthetic vegetation in remote-sensing-based estimates of carbon flux in wetlands

Monitoring productivity in coastal wetlands is important due to their high carbon sequestration rates and potential role in climate change mitigation. We tested agricultural- and forest-based methods for estimating the fraction of absorbed photosynthetically active radiation (f _APAR), a key parameter for modelling gross primary productivity (GPP), in a restored, managed wetland with a dense litter layer of non-photosynthetic vegetation, and we compared the difference in canopy light transmission between a tidally influenced wetland and the managed wetland. The presence of litter reduced correlations between spectral vegetation indices and f _APAR. In the managed wetland, a two-band vegetation index incorporating simulated World View-2 or Hyperion green and near-infrared bands, collected with a field spectroradiometer, significantly correlated with f _APAR only when measured above the litter layer, not at the ground where measurements typically occur. Measures of GPP in these systems are difficult to capture via remote sensing, and require an investment of sampling effort, practical methods for measuring green leaf area and accounting for background effects of litter and water.     

Cadmium tolerance and accumulation characteristics of wetland emergent plants under hydroponic conditions

For the purpose of screening a potential Cd-hyperaccumulator for Cd-contaminated soil in paddy fields, four kinds of wetland emergent plants (Iris sibirica L., Acorus calamus L., Typha orientalis Presl and Cyperus alternifolius L.) were investigated for their cadmium tolerance and accumulation characteristics under hydroponic conditions. The physiological responses of plants, Cd concentration in tissues, Cd accumulation, bioaccumulation factor (BCF) and translocation factor (TCF) were investigated to evaluate the abilities of wetland emergent plants to absorb and accumulate Cd. In comparison with the other selected emergent plants, Iris sibirica L. has the strongest Cd-tolerance for the absence of Cd toxic symptoms and a Cd concentration as high as 127.3 mg kg⁻¹ in shoots. Due to its large biomass, the Cd accumulation could reach up to 9.4 mg per plant in roots and 5.7 mg per plant in shoots, respectively. Iris sibirica L. possesses the highest TCF, and its BCF for Cd increased with increasing concentration of spiked Cd in the hydroponic solutions. The results indicate that Iris sibirica L. is a potential Cd-hyperaccumulator that may have a strong capacity for extracting Cd from Cd-contaminated paddy soils.

Four kinds of wetland emergent plants (Iris sibirica L., Acorus calamus L., Typha orientalis Presl and Cyperus alternifolius L.) were investigated for their cadmium tolerance and accumulation characteristics under hydroponic conditions.     

An improved water-depth correction algorithm for seagrass mapping using hyperspectral data

Remote detection of seagrass has been limited because of numerous factors including the influence of the water column, which interferes with reflected signals from the seafloor. In a previously published study, a water-depth correction algorithm was developed to improve the detection of underwater vegetation spectral signals. The algorithm successfully corrected laboratory-measured submerged vegetation spectra for water effects, but the water absorption coefficients, derived from the data collected over a white surface, tended to underestimate the actual water absorption when applied to hyperspectral image data. The experimental conditions were modified to reduce the errors associated with the effects of enhanced multi-path scattering, to improve the algorithm using the new empirical data and to apply the algorithm to an airborne hyperspectral image data obtained over Halodule wrightii seagrass beds at Grand Bay National Estuarine Research Reserve, Mississippi, USA. The water absorption and scattering factors (A _w and R _w) for a water depth of 40 cm (the local water thickness above the seagrass canopy measured in the field) were applied to the image data to obtain the reflectance that is attributed to the water bottom surface including bare sand and seagrass beds. The contrast between the dark Halodule patches and the bright sand increased in the bands between 500 and 800 nm after the correction. The correction algorithm also increased the normalized difference vegetation index (NDVI) values for the seagrass pixels by restoring the upwelling signal in the near-infrared.     

Bias in lidar-based canopy gap fraction estimates

Leaf area index and canopy gap fraction (GF) provide important information to forest managers regarding the ecological functioning and productivity of forest resources. Traditional measurements such as those obtained from hemispherical photography (HP) measure solar irradiation, penetrating the forest canopy, but do not provide information regarding the three-dimensional canopy structure. Terrestrial laser scanning (TLS) is an active sensor technology able to describe structural forest attributes by measuring interceptions of emitted laser pulses with the canopy and is able to record the spatial distribution of the foliage in three dimensions. However, due to the beam area of the laser, interceptions are detected more frequently than using conventional HP, and GF is typically underestimated. This study investigates the effects of laser beam area on the retrieval of GF by using morphological image processing to describe estimation bias as a function of canopy perimeters. The results show that, using canopy perimeter, improvements in correlation between HP and TLS can be achieved with an increase in the coefficient of determination R ² up to 28% (from an original R ² of 0.66 to an adjusted R ² of 0.85).     

Estimating dry matter content of fresh leaves from the residuals between leaf and water reflectance

At about 1720 nm wavelength, there is an absorption feature of leaf dry matter based on a C?H stretch overtone, which is difficult to detect in fresh green leaves because of the absorption spectrum of liquid water. We applied a method originally proposed in Remote Sensing of Environment by B.-C. Gao and A.F.H. Goetz (Estimating dry matter content of fresh leaves from the residuals between leaf and water reflectance, pp. 137–145, 1994) that used linear regression between the natural logarithm of leaf spectral reflectance and the specific absorption coefficient of liquid water over wavelengths from 1500 to 1800 nm to calculate an expected reflectance spectrum. The residual difference between the measured and expected leaf reflectance spectra enhanced the absorption feature of dry matter. The absorption feature was quantified using the normalized dry matter index (NDMI) based on the contrast between the residual leaf reflectance at 1649 and 1722 nm. NDMI was linearly related to leaf dry matter content (g cm^?2) for data obtained in the field (coefficient of determination r ² = 0.636) and for the leaf optical properties experiment (LOPEX; r ² = 0.684). Lignin and cellulose contents were also measured during LOPEX, but NDMI was weakly correlated to these variables, indicating NDMI was sensitive to all leaf biochemical constituents. Estimation of dry matter content combined with estimates of water content will facilitate the calculation of the fuel moisture content for prediction of wildfire.     

A simple spectro-goniometer for collection of multiple view angle reflectance factors

Field spectroradiometers are widely used for environmental applications where data describing visible and near-infrared reflectance factors are of interest. Recent developments in spaceborne and airborne instruments with multiple view angle (MVA) capabilities have resulted in a demand for ground measurements to support these missions. Lightweight portable spectroradiometers offer an appropriate means of collecting MVA spectral reflectance factor data because they are more easily manoeuvrable than other spectroradiometers, but their physical capabilities have not yet been explored in this context. This letter presents the results of a focused experiment aimed at evaluating the field capabilities of a miniaturized Ocean Optics instrument in MVA settings for soil surface roughness applications. MVA hemispherical-conical reflectance factors were collected in situ from soil surfaces whose roughness was determined using a laser profiling survey. The results showed a significant negative relationship (R ²?=?0.74; p < 0.01) between directional reflectance factors measured at 870 nm in the forward-scattering region and a soil structural measure derived from laser profiling data. This corroborates the results of other published studies and suggests that Ocean Optics instruments can be used to support hyperspectral MVA investigations.     

Lake water surface mapping in the Tibetan Plateau using the MODIS MOD09Q1 product

The Tibetan Plateau (TP) has the largest number of inland lakes with the highest elevation on the planet. Mapping the distribution of lake water in space and time is crucial for scientific research of interactions among the regional cryosphere, hydrosphere, and atmosphere. In this study, a lake water surface mapping algorithm is developed for Moderate Resolution Imaging Spectroradiometer (MODIS) MOD09Q1 surface reflectance images, which is used to produce the 8-day lake water surface data set (lake water surface area larger than 1 km²) of theTP (Qinghai–Tibet Plateau) for the period of 2000–2012. The accuracy analysis indicate that compared with water surface data of the 134 sample lakes extracted from the 30 m Landsat Thematic Mapper (TM) images, the average overall accuracy of the results is 91.81% with average commission and omission error of 3.26% and 5.38%; the results also show strong linear (the coefficient of determination R² is 0.9991) correlation with the global MODIS water mask data set with overall accuracy of 86.30%; and the lake area difference between the Second National Lake Survey and this study is only 4.74%, respectively. This study provides reliable data set for the lake change research of theTP in the recent decade.     

Data-driven correction for light attenuation in shallow waters

Application of remote sensing to aquatic habitats is generally complex, due to the presence of optically active components that absorb or scatter light. According to the Beer–Lambert law, as electromagnetic energy travels through a medium, it is attenuated at an exponential rate determined by the physical and chemical properties of the medium. Determination of attenuation coefficients is complex for optically shallow inland and coastal waters due to bottom reflection and/or multi-path scattering. To better understand light attenuation in shallow waters, mathematical relationships between water depth and vertical attenuation coefficient (K _d) were derived using experimental spectral data (400–900 nm). The model was used to estimate K _d values at water depths that were beyond and in between the experimentally measured points (5–60 cm). The depth- and wavelength-dependent K _d values were used to correct the measured spectral reflectance for the corresponding light attenuation to restore the bottom reflectance signals. After the correction, the spectral reflectance patterns measured above submerged aquatic plants appeared to include a restored signal from vegetation especially in the near-infrared (NIR) region. Comparison between the modelled spectra and the empirically measured spectra suggest that the proposed approach produced an effective model for light attenuation in optically shallow waters.     

Feasibility of using mobile phone to estimate forest Leaf Area Index: a case study in Yunnan Pine

Remotely-sensed Leaf Area Index (LAI) is vital to describe the vegetation canopy and assess plant growth condition and healthy status. However, sufficient instant ground LAI samples are pre-required for calibration or validation, which is generally difficult to collect. We proposed a method, LAI-Mobile, to use mobile phones with low-cost fisheye lens to take fisheye photos and to invert LAIs, which may be popularized for ordinary people to generate big volume of LAI sample data. The feasibility of LAI-Mobile was tested by comparing with LAI?2200 and GF-1 satellite data (GF = high resolution) in a pest-invaded Yunnan pine forest area in Yunnan province of China. Results show significant correlation between LAI-2200 and LAI-Mobile data for forest plots with coefficient of determination (R²) = 0.706 and Root Mean Square Error (RSME) = 0.241, and GF-1 satellite images (R² = 0.659 and RMSE = 0.268). The linear regression shows a good agreement between the Moderate Resolution Imaging Spectroradiometer (MODIS) LAI product and the inverted GF-1 LAI, with R² = 0.649, RMSE = 0.795. Despite larger uncertainty for single fisheye image than LAI-2200, LAI-mobile can provide fast and convenient method to collect large amount of LAI, which will support remote sensing inversion of LAI at large scale.     

Mapping tree stress associated with urban pollution using the WorldView-2 Red Edge band

A small urban park in Yerevan, Armenia, was studied using biogechemical analysis of the tree canopy, field spectral reflectance measurements of tree leaves, simulated WorldView-2 multispectral data generated from the leaf spectra, and two summer images of real WorldView-2 data. The tree canopy of the park is dominated by two trees, Robinia pseudoacacia L. (locust) and Fraxinus excelsior L. (ash). The Highest values of lead, nickel, molybdenum, copper and zinc were found in leaves harvested from trees adjacent to the streets, whereas most of the lowest values for those metals were found in the interior of the park. A t-test of the field spectral measurements indicated that the green and red edge spectral reflectance of leaves from trees near the streets was significantly higher than that of leaves of trees in the interior (p < 0.05). However, in simulated WorldView-2 multispectral data, the street and interior leaves were only statistically separable in band 6 (Red Edge) raw data and hyperspherical direction cosine (HSDC) normalized band 6 data. HSDC-normalized band 6 digital numbers from real WorldView-2 data of 16 June and 9 August 2011 from trees adjacent to the streets were statistically higher than the interior locations for both dates. Maps of anomalously high HSDC-normalized band 6 values show a concentration on the park edges, suggesting vehicle pollution may indeed be the cause of the observed patterns.     

Predicting distribution of microphytobenthos abundance on a reef platform by combining in situ underwater spectrometry and pigment analysis

An approach is presented to predict microphytobenthos (MPB) abundance at Heron Reef, Australia using a regression model between chlorophyll-a (chl-a) concentration and field measured spectral reflectance integrated to the blue spectral band of the Worldview-2 (WV-2) sensor. In situ underwater spectral reflectance was measured from benthic sediments and co-located sediment cores were collected to estimate their chl-a concentration. Chl-a features were visually identified in spectral reflectance curves by derivative analysis. Pigment composition analysis of benthic sediments enabled the presence of a relatively homogeneous and living MPB community to be inferred. Field spectral reflectance and chl-a concentration showed a negative correlation in blue and red wavelengths. Regression models between field spectral reflectance resampled to the WV-2 sensor’s bandwidths and chl-a concentration showed the highest coefficient of determination in the blue spectral band of the sensor. This ‘blue band’ model can be further applied to individual pixels of WV-2 images to produce spatially explicit and continuous maps of chl-a concentration, and hence MPB abundance, in Heron Reef.     

Assessment of spectral indices for cover estimation of senescent vegetation

Quantification of dry plant matter (crop residue, senesced foliage, non-photosynthetic vegetation, or plant litter) surface cover (f _R) is important for assessing agricultural tillage practices, carbon sequestration, rangeland health, or brush fire hazards. The Cellulose Absorption Index (CAI) and the Shortwave Infrared Normalized Difference Residue Index (SINDRI) are two spectral indices that can remotely estimate f _R. CAI and SINDRI utilize three and two spectral bands, respectively, so SINDRI is expected to be less expensive to implement in future satellite sensors. We assessed the contrast of CAI and SINDRI with respect to soil reflectance spectra. Estimating f _R with CAI is possible for all soils. However, a number of soil samples had positive SINDRI values due to various soil minerals, such as gibbsite and antigorite, which would be interpreted as high f _R, and could limit its usefulness in some areas. Therefore, SINDRI is less applicable for estimating f _R, even with reduced implementation costs.     

Performance of two neural network models in bathymetry

Neural networks are widely used as predictors in several fields of applications, such as prediction of shallow water depth. The purpose of this study is to investigate the performance of two artificial neural networks models as potential methods in bathymetry. A comparison approach is used to evaluate network models, the regression tree and an inversion model. The high-resolution IKONOS and moderate-resolution Landsat satellite images serve as the case studies, and results based on the root mean square errors and coefficient of determination (R²) show that artificial neural networks outperform the inversion model and the regression tree.