Above ground biomass estimation in an African tropical forest with lidar and hyperspectral data

The estimation of above ground biomass in forests is critical for carbon cycle modeling and climate change mitigation programs. Small footprint lidar provides accurate biomass estimates, but its application in tropical forests has been limited, particularly in Africa. Hyperspectral data record canopy spectral information that is potentially related to forest biomass. To assess lidar ability to retrieve biomass in an African forest and the usefulness of including hyperspectral information, we modeled biomass using small footprint lidar metrics as well as airborne hyperspectral bands and derived vegetation indexes. Partial Least Square Regression (PLSR) was adopted to cope with multiple inputs and multicollinearity issues; the Variable of Importance in the Projection was calculated to evaluate importance of individual predictors for biomass. Our findings showed that the integration of hyperspectral bands (R² = 0.70) improved the model based on lidar alone (R² = 0.64), this encouraging result call for additional research to clarify the possible role of hyperspectral data in tropical regions. Replacing the hyperspectral bands with vegetation indexes resulted in a smaller improvement (R² = 0.67). Hyperspectral bands had limited predictive power (R² = 0.36) when used alone. This analysis proves the efficiency of using PLSR with small-footprint lidar and high resolution hyperspectral data in tropical forests for biomass estimation. Results also suggest that high quality ground truth data is crucial for lidar-based AGB estimates in tropical African forests, especially if airborne lidar is used as an intermediate step of upscaling field-measured AGB to a larger area.     

Commercial tree species discrimination using airborne AISA Eagle hyperspectral imagery and partial least squares discriminant analysis (PLS-DA) in KwaZulu–Natal,South Africa

Discriminating commercial tree species using hyperspectral remote sensing techniques is critical in monitoring the spatial distributions and compositions of commercial forests. However, issues related to data dimensionality and multicollinearity limit the successful application of the technology. The aim of this study was to examine the utility of the partial least squares discriminant analysis (PLS-DA) technique in accurately classifying six exotic commercial forest species (Eucalyptus grandis, Eucalyptus nitens, Eucalyptus smithii, Pinus patula, Pinus elliotii and Acacia mearnsii) using airborne AISA Eagle hyperspectral imagery (393–900 nm). Additionally, the variable importance in the projection (VIP) method was used to identify subsets of bands that could successfully discriminate the forest species. Results indicated that the PLS-DA model that used all the AISA Eagle bands (n = 230) produced an overall accuracy of 80.61% and a kappa value of 0.77, with user’s and producer’s accuracies ranging from 50% to 100%. In comparison, incorporating the optimal subset of VIP selected wavebands (n = 78) in the PLS-DA model resulted in an improved overall accuracy of 88.78% and a kappa value of 0.87, with user’s and producer’s accuracies ranging from 70% to 100%. Bands located predominantly within the visible region of the electromagnetic spectrum (393–723 nm) showed the most capability in terms of discriminating between the six commercial forest species. Overall, the research has demonstrated the potential of using PLS-DA for reducing the dimensionality of hyperspectral datasets as well as determining the optimal subset of bands to produce the highest classification accuracies.     

Vegetation chlorophyll estimates in the Amazon from multi-angle MODIS observations and canopy reflectance model

As a preparatory study for future hyperspectral missions that can measure canopy chemistry, we introduce a novel approach to investigate whether multi-angle Moderate Resolution Imaging Spectroradiometer (MODIS) data can be used to generate a preliminary database with long-term estimates of chlorophyll. MODIS monthly chlorophyll estimates between 2000 and 2015, derived from a fully coupled canopy reflectance model (ProSAIL), were inspected for consistency with eddy covariance fluxes, tower-based hyperspectral images and chlorophyll measurements. MODIS chlorophyll estimates from the inverse model showed strong seasonal variations across two flux-tower sites in central and eastern Amazon. Marked increases in chlorophyll concentrations were observed during the early dry season. Remotely sensed chlorophyll concentrations were correlated to field measurements (r² = 0.73 and r² = 0.98) but the data deviated from the 1:1 line with root mean square errors (RMSE) ranging from 0.355 μg cm⁻² (Tapajós tower) to 0.470 μg cm⁻² (Manaus tower). The chlorophyll estimates were consistent with flux tower measurements of photosynthetically active radiation (PAR) and net ecosystem productivity (NEP). We also applied ProSAIL to mono-angle hyperspectral observations from a camera installed on a tower to scale modeled chlorophyll pigments to MODIS observations (r² = 0.73). Chlorophyll pigment concentrations (Chl_A+B) were correlated to changes in the amount of young and mature leaf area per month (0.59 ≤  r² ≤  0.64). Increases in MODIS observed Chl_A+B were preceded by increased PAR during the dry season (0.61 ≤ r² ≤  0.62) and followed by changes in net carbon uptake. We conclude that, at these two sites, changes in LAI, coupled with changes in leaf chlorophyll, are comparable with seasonality of plant productivity. Our results allowed the preliminary development of a 15-year time series of chlorophyll estimates over the Amazon to support canopy chemistry studies using future hyperspectral sensors.     

Comparison of MODIS-based models for retrieving suspended particulate matter concentrations in Poyang Lake,China

Suspended particulate matter (SPM) is a key parameter describing water quality, and developing the retrieval model of SPM concentration (C_SPM) is fundamental for obtaining the spatiotemporal information of C_SPM and further for understanding, managing and protecting aquatic ecosystems. This study aimed to compare moderate resolution imaging spectroradiometer (MODIS)-based C_SPM retrieval models in order to find the optimal model for improving the C_SPM estimation in Poyang Lake. The C_SPM measurements on 27 September 2007 and their coincident MODIS Terra image were used to calibrate retrieval models with the least-squares technique. The C_SPM measurements on 31 August 2012 and the MODIS Terra image on 30 August 2012 were applied to validate the calibrated models, and the correlation coefficient (r) between the measured and estimated C_SPM values, the root mean square error (RMSE) and relative root mean square error (RRMSE) of estimation as well as the model bias evaluation result were compared to determine the optimal model for estimating the C_SPM values of Poyang Lake from MODIS images. Model calibration showed that, after two samples were removed, the exponential models of blue, green and red band, the linear model of infrared band, the cubic model of red band as well as the exponential model of red minus infrared band explained about 92%, 88%, 90%, 89%, 90% and 76% of the variation of C_SPM, respectively; while model validation indicated that, after removing two samples, the exponential models of blue and green band got biased C_SPM estimations, the agreement between the measured and estimated C_SPM values was not very high (r = <0.8) for the models with single red and infrared band, and the exponential model of red minus infrared band got the best result among all calibrated models (r = 0.87, RMSE = 22.1 mg/l, RRMSE = 52.8%). We concluded that the exponential model of red minus infrared band obtained stable C_SPM estimation and was the optimal model for C_SPM estimation in this study, and more independent datasets should be obtained to further validate our finding for improving the C_SPM estimation in Poyang Lake.     

Improving the prediction of arsenic contents in agricultural soils by combining the reflectance spectroscopy of soils and rice plants

Visible and near-infrared reflectance spectroscopy provides a beneficial tool for investigating soil heavy metal contamination. This study aimed to investigate mechanisms of soil arsenic prediction using laboratory based soil and leaf spectra, compare the prediction of arsenic content using soil spectra with that using rice plant spectra, and determine whether the combination of both could improve the prediction of soil arsenic content. A total of 100 samples were collected and the reflectance spectra of soils and rice plants were measured using a FieldSpec3 portable spectroradiometer (350–2500 nm). After eliminating spectral outliers, the reflectance spectra were divided into calibration (n = 62) and validation (n = 32) data sets using the Kennard-Stone algorithm. Genetic algorithm (GA) was used to select useful spectral variables for soil arsenic prediction. Thereafter, the GA-selected spectral variables of the soil and leaf spectra were individually and jointly employed to calibrate the partial least squares regression (PLSR) models using the calibration data set. The regression models were validated and compared using independent validation data set. Furthermore, the correlation coefficients of soil arsenic against soil organic matter, leaf arsenic and leaf chlorophyll were calculated, and the important wavelengths for PLSR modeling were extracted. Results showed that arsenic prediction using the leaf spectra (coefficient of determination in validation, R_v² = 0.54; root mean square error in validation, RMSE_v = 12.99 mg kg⁻¹; and residual prediction deviation in validation, RPD_v = 1.35) was slightly better than using the soil spectra (R_v² = 0.42, RMSE_v = 13.35 mg kg⁻¹, and RPD_v = 1.31). However, results also showed that the combinational use of soil and leaf spectra resulted in higher arsenic prediction (R_v² = 0.63, RMSE_v = 11.94 mg kg⁻¹, RPD_v = 1.47) compared with either soil or leaf spectra alone. Soil spectral bands near 480, 600, 670, 810, 1980, 2050 and 2290 nm, leaf spectral bands near 700, 890 and 900 nm in PLSR models were important wavelengths for soil arsenic prediction. Moreover, soil arsenic showed significantly positive correlations with soil organic matter (r = 0.62, p < 0.01) and leaf arsenic (r = 0.77, p < 0.01), and a significantly negative correlation with leaf chlorophyll (r = −0.67, p < 0.01). The results showed that the prediction of arsenic contents using soil and leaf spectra may be based on their relationships with soil organic matter and leaf chlorophyll contents, respectively. Although RPD of 1.47 was below the recommended RPD of >2 for soil analysis, arsenic prediction in agricultural soils can be improved by combining the leaf and soil spectra.     

Remotely detecting canopy nitrogen concentration and uptake of paddy rice in the Northeast China Plain

The influence of morphophysiological variation at different growth stages on the performance of vegetation indices for estimating plant N status has been confirmed. However, the underlying mechanisms explaining how this variation impacts hyperspectral measures and canopy N status are poorly understood. In this study, four field experiments involving different N rates were conducted to optimize the selection of sensitive bands and evaluate their performance for modeling canopy N status of rice at various growth stages in 2007 and 2008. The results indicate that growth stages negatively affect hyperspectral indices in different ways in modeling leaf N concentration (LNC), plant N concentration (PNC) and plant N uptake (PNU). Published hyperspectral indices showed serious limitations in estimating LNC, PNC and PNU. The newly proposed best 2-band indices significantly improved the accuracy for modeling PNU (R² = 0.75–0.85) by using the lambda by lambda band-optimized algorithm. However, the newly proposed 2-band indices still have limitations in modeling LNC and PNC because the use of only 2-band indices is not fully adequate to provide the maximum N-related information. The optimum multiple narrow band reflectance (OMNBR) models significantly increase the accuracy for estimating the LNC (R² = 0.67–0.71) and PNC (R² = 0.57–0.78) with six bands. Results suggest the combinations of center of red-edge (735 nm) with longer red-edge bands (730–760 nm) are very efficient for estimating PNC after heading, whereas the combinations of blue with green bands are more efficient for modeling PNC across all stages. The center of red-edge (730–735 nm) paired with early NIR bands (775–808 nm) are predominant in estimating PNU before heading, whereas the longer red-edge (750 nm) paired with the center of “NIR shoulder” (840–850 nm) are dominant in estimating PNU after heading and across all stages. The OMNBR models have the advantage of modeling canopy N status for the entire growth period. However, the best 2-band indices are much easier to use. Alternatively, it is also possible to use the best 2-band indices to monitor PNU before heading and PNC after heading. This study systematically explains the influences of N dilution effect on hyperspectral band combinations in relating to the different N variables and further recommends the best band combinations which may provide an insight for developing new hyperspectral vegetation indices.     

Retrieving aboveground biomass of wetland Phragmites australis (common reed) using a combination of airborne discrete-return LiDAR and hyperspectral data

Wetland biomass is essential for monitoring the stability and productivity of wetland ecosystems. Conventional field methods to measure or estimate wetland biomass are accurate and reliable, but expensive, time consuming and labor intensive. This research explored the potential for estimating wetland reed biomass using a combination of airborne discrete-return Light Detection and Ranging (LiDAR) and hyperspectral data. To derive the optimal predictor variables of reed biomass, a range of LiDAR and hyperspectral metrics at different spatial scales were regressed against the field-observed biomasses. The results showed that the LiDAR-derived H_p99 (99th percentile of the LiDAR height) and hyperspectral-calculated modified soil-adjusted vegetation index (MSAVI) were the best metrics for estimating reed biomass using the single regression model. Although the LiDAR data yielded a higher estimation accuracy compared to the hyperspectral data, the combination of LiDAR and hyperspectral data produced a more accurate prediction model for reed biomass (R² = 0.648, RMSE = 167.546 g/m², RMSE_r = 20.71%) than LiDAR data alone. Thus, combining LiDAR data with hyperspectral data has a great potential for improving the accuracy of aboveground biomass estimation.     

Spectral detection of stress-related pigments in salt-lake succulent halophytic shrubs

The spectral detection of vegetation pigment concentrations has a high potential value, but it is still underdeveloped, especially for pigments other than chlorophylls. In this study, the seasonal pigment dynamics of two Tecticornia species (samphires; halophytic shrubs) from north-western Australia were correlated with spectral indices that best document the pigment changes over time. Pigment dynamics were assessed by analysing betacyanin, chlorophyll and carotenoid concentrations at plant level and by measuring reflectance at contrasting seasonal dates. Plant reflectance was used to define a new reflectance index that was most sensitive to the seasonal shifts in Tecticornia pigment concentrations. The two Tecticornia species turned from green to red-pinkish for the period March–August 2012 when betacyanins increased almost nine times in both species. Chlorophyll levels showed the opposite pattern to that of betacyanins, whereas carotenoid levels were relatively stable. Normalised difference indices correlated well with betacyanin (r = 0.805, using bands at 600 and 620 nm) and chlorophyll (r = 0.809, using bands at 737 and 726 nm). Using knowledge of chlorophyll concentrations slightly improved the ability of the spectral index to predict betacyanin concentration (r = 0.822 at bands 606 and 620 nm, in the case of chemically determined chlorophyll, r = 0.809 when using remotely sensed chlorophyll). Our results suggest that this new spectral index can reliably detect changes in betacyanin concentrations in vegetation, with potential applications in ecological studies and environmental impact monitoring.     

Understanding the optical responses of leaf nitrogen in Mediterranean Holm oak (Quercus ilex) using field spectroscopy

The direct estimation of nitrogen (N) in fresh vegetation is challenging due to its weak influence on leaf reflectance and the overlaps with absorption features of other compounds. Different empirical models relate in this work leaf nitrogen concentration ([N]_Leaf) on Holm oak to leaf reflectance as well as derived spectral indices such as normalized difference indices (NDIs), the three bands indices (TBIs) and indices previously used to predict leaf N and chlorophyll. The models were calibrated and assessed their accuracy, robustness and the strength of relationship when other biochemicals were considered. Red edge was the spectral region most strongly correlated with [N]_Leaf, whereas most of the published spectral indexes did not provide accurate estimations. NDIs and TBIs based models could achieve robust and acceptable accuracies (TBI_{1310,1720,730}: R² = 0.76, [0.64,0.86]; RMSE (%) = 9.36, [7.04,12.83]). These models sometimes included indices with bands close to absorption features of N bonds or nitrogenous compounds, but also of other biochemicals. Models were independently and inter-annually validated using the bootstrap method, which allowed discarding those models non-robust across different years. Partial correlation analysis revealed that spectral estimators did not strongly respond to [N]_Leaf but to other leaf variables such as chlorophyll and water, even if bands close to absorption features of N bonds or compounds were present in the models.     

Modelling the standing timber volume of Baden-Württemberg—A large-scale approach using a fusion of Landsat,airborne LiDAR and National Forest Inventory data

Remote sensing-based timber volume estimation is key for modelling the regional potential, accessibility and price of lignocellulosic raw material for an emerging bioeconomy. We used a unique wall-to-wall airborne LiDAR dataset and Landsat 7 satellite images in combination with terrestrial inventory data derived from the National Forest Inventory (NFI), and applied generalized additive models (GAM) to estimate spatially explicit timber distribution and volume in forested areas. Since the NFI data showed an underlying structure regarding size and ownership, we additionally constructed a socio-economic predictor to enhance the accuracy of the analysis. Furthermore, we balanced the training dataset with a bootstrap method to achieve unbiased regression weights for interpolating timber volume. Finally, we compared and discussed the model performance of the original approach (r² = 0.56, NRMSE = 9.65%), the approach with balanced training data (r² = 0.69, NRMSE = 12.43%) and the final approach with balanced training data and the additional socio-economic predictor (r² = 0.72, NRMSE = 12.17%). The results demonstrate the usefulness of remote sensing techniques for mapping timber volume for a future lignocellulose-based bioeconomy.     

Assessment of RapidEye vegetation indices for estimation of leaf area index and biomass in corn and soybean crops

Leaf area index (LAI) and biomass are important indicators of crop development and the availability of this information during the growing season can support farmer decision making processes. This study demonstrates the applicability of RapidEye multi-spectral data for estimation of LAI and biomass of two crop types (corn and soybean) with different canopy structure, leaf structure and photosynthetic pathways. The advantages of Rapid Eye in terms of increased temporal resolution (∼daily), high spatial resolution (∼5 m) and enhanced spectral information (includes red-edge band) are explored as an individual sensor and as part of a multi-sensor constellation. Seven vegetation indices based on combinations of reflectance in green, red, red-edge and near infrared bands were derived from RapidEye imagery between 2011 and 2013. LAI and biomass data were collected during the same period for calibration and validation of the relationships between vegetation indices and LAI and dry above-ground biomass. Most indices showed sensitivity to LAI from emergence to 8 m²/m². The normalized difference vegetation index (NDVI), the red-edge NDVI and the green NDVI were insensitive to crop type and had coefficients of variations (CV) ranging between 19 and 27%; and coefficients of determination ranging between 86 and 88%. The NDVI performed best for the estimation of dry leaf biomass (CV = 27% and r² = 090) and was also insensitive to crop type. The red-edge indices did not show any significant improvement in LAI and biomass estimation over traditional multispectral indices. Cumulative vegetation indices showed strong performance for estimation of total dry above-ground biomass, especially for corn (CV ≤ 20%). This study demonstrated that continuous crop LAI monitoring over time and space at the field level can be achieved using a combination of RapidEye, Landsat and SPOT data and sensor-dependant best-fit functions. This approach eliminates/reduces the need for reflectance resampling, VIs inter-calibration and spatial resampling.     

Retrieval of tea polyphenol at leaf level using spectral transformation and multi-variate statistical approach

In the present study, field based hyperspectral data was used to estimate the tea (Camellia sinensis L.) polyphenol at Deha Tea garden of Assam state, India. Leaf reflectance spectra were first filtered for noise and then transformed into normalized and first derivative reflectance for further analysis. Stepwise discriminant analysis was carried out to select sensitive bands for a range of polyphenol concentration by minimizing the effects of other factors such as age of the bushes and management practices. The wavelengths at 358, 369, 484, 845, 916, 1387, 1420, 1435, 1621 and 2294 nm were identified as sensitive to tea polyphenol, among which 2294 nm was found to be the most recurring band. The noise removed selected bands, their transformed derivatives and principal components were regressed with the tea polyphenol using univariate and multi-variate analysis. In univariate analysis the correlation was very poor with RMSE more than 3.0. A significant improvement in R² values were observed when multivariate analyses like stepwise multiple linear regression (SMLR) and partial least square regression (PLSR) was carried out. The PLSR of first derivative reflectance was most accurate (R² = 0.81 and RMSE = 1.39 mg g⁻¹) among all the uni- and multivariate analysis for predicting the polyphenol of fresh tea leaves.     

Non-linear partial least square regression increases the estimation accuracy of grass nitrogen and phosphorus using in situ hyperspectral and environmental data

Grass nitrogen (N) and phosphorus (P) concentrations are direct indicators of rangeland quality and provide imperative information for sound management of wildlife and livestock. It is challenging to estimate grass N and P concentrations using remote sensing in the savanna ecosystems. These areas are diverse and heterogeneous in soil and plant moisture, soil nutrients, grazing pressures, and human activities. The objective of the study is to test the performance of non-linear partial least squares regression (PLSR) for predicting grass N and P concentrations through integrating in situ hyperspectral remote sensing and environmental variables (climatic, edaphic and topographic). Data were collected along a land use gradient in the greater Kruger National Park region. The data consisted of: (i) in situ-measured hyperspectral spectra, (ii) environmental variables and measured grass N and P concentrations. The hyperspectral variables included published starch, N and protein spectral absorption features, red edge position, narrow-band indices such as simple ratio (SR) and normalized difference vegetation index (NDVI). The results of the non-linear PLSR were compared to those of conventional linear PLSR. Using non-linear PLSR, integrating in situ hyperspectral and environmental variables yielded the highest grass N and P estimation accuracy (R² = 0.81, root mean square error (RMSE) = 0.08, and R² = 0.80, RMSE = 0.03, respectively) as compared to using remote sensing variables only, and conventional PLSR. The study demonstrates the importance of an integrated modeling approach for estimating grass quality which is a crucial effort towards effective management and planning of protected and communal savanna ecosystems.     

Impact of sub-pixel heterogeneity on modelled brightness temperature for an agricultural region

Knowledge of sub-pixel heterogeneity, particularly at the passive microwave scale, can improve the brightness temperature (and ultimately the soil moisture) estimation. However, the impact of surface heterogeneity (in terms of soil moisture, soil temperature and vegetation water content) on brightness temperature in an agricultural setting is relatively unknown. The Soil Moisture Active Passive Validation Experiment 2012 (SMAPVEX12) provided an opportunity to evaluate sub-pixel heterogeneity at the scale of a Soil Moisture Ocean Salinity (SMOS) or the Soil Moisture Active Passive (SMAP) radiometer footprint using field measured data. The first objective of this study was to determine if accounting for surface heterogeneity reduced the error between estimated brightness temperature (T_b) and T_b measured by SMOS. It was found that when accounting for variation in surface soil moisture, temperature and vegetation water content within the pixel footprint, the error between the modelled T_b and the measured T_b was less than if a homogeneous pixel were modelled. The correlation between the surface parameters and the error associated with not accounting for surface heterogeneity were investigated. It was found that there was low to moderate correlation between the error and the coefficient of variance associated with the measured soil moisture, soil temperature and vegetation volumetric water content during the field campaign. However, it was found that the correlations changed depending on the stage of vegetation growth and the amount of time following a precipitation event. At the start of the field campaign (following a precipitation event), there was strong correlation between the error and all three surface parameters (r ≥ 0.75). Following a precipitation event close to the middle of the field campaign (during which there was rapid growth in vegetation), there was strong correlation between the error and the variability in vegetation water content (r = 0.89), moderate correlation with soil moisture (r = 0.61) and low correlation with soil temperature (r = 0.26).     

Modeling soil parameters using hyperspectral image reflectance in subtropical coastal wetlands

Developing spectral models of soil properties is an important frontier in remote sensing and soil science. Several studies have focused on modeling soil properties such as total pools of soil organic matter and carbon in bare soils. We extended this effort to model soil parameters in areas densely covered with coastal vegetation. Moreover, we investigated soil properties indicative of soil functions such as nutrient and organic matter turnover and storage. These properties include the partitioning of mineral and organic soil between particulate (>53 μm) and fine size classes, and the partitioning of soil carbon and nitrogen pools between stable and labile fractions. Soil samples were obtained from Avicennia germinans mangrove forest and Juncus roemerianus salt marsh plots on the west coast of central Florida. Spectra corresponding to field plot locations from Hyperion hyperspectral image were extracted and analyzed. The spectral information was regressed against the soil variables to determine the best single bands and optimal band combinations for the simple ratio (SR) and normalized difference index (NDI) indices. The regression analysis yielded levels of correlation for soil variables with R² values ranging from 0.21 to 0.47 for best individual bands, 0.28 to 0.81 for two-band indices, and 0.53 to 0.96 for partial least-squares (PLS) regressions for the Hyperion image data. Spectral models using Hyperion data adequately (RPD > 1.4) predicted particulate organic matter (POM), silt + clay, labile carbon (C), and labile nitrogen (N) (where RPD = ratio of standard deviation to root mean square error of cross-validation [RMSECV]). The SR (0.53 μm, 2.11 μm) model of labile N with R² = 0.81, RMSECV= 0.28, and RPD = 1.94 produced the best results in this study. Our results provide optimism that remote-sensing spectral models can successfully predict soil properties indicative of ecosystem nutrient and organic matter turnover and storage, and do so in areas with dense canopy cover.     

Detecting Sirex noctilio grey-attacked and lightning-struck pine trees using airborne hyperspectral data,random forest and support vector machines classifiers

The visual progression of sirex (Sirex noctilio) infestation symptoms has been categorized into three distinct infestation phases, namely the green, red and grey stages. The grey stage is the final stage which leads to almost complete defoliation resulting in dead standing trees or snags. Dead standing pine trees however, could also be due to the lightning damage. Hence, the objective of the present study was to distinguish amongst healthy, sirex grey-attacked and lightning-damaged pine trees using AISA Eagle hyperspectral data, random forest (RF) and support vector machines (SVM) classifiers. Our study also presents an opportunity to look at the possibility of separating amongst the previously mentioned pine trees damage classes and other landscape classes on the study area. The results of the present study revealed the robustness of the two machine learning classifiers with an overall accuracy of 74.50% (total disagreement = 26%) for RF and 73.50% (total disagreement = 27%) for SVM using all the remaining AISA Eagle spectral bands after removing the noisy ones. When the most useful spectral bands as measured by RF were exploited, the overall accuracy was considerably improved; 78% (total disagreement = 22%) for RF and 76.50% (total disagreement = 24%) for SVM. There was no significant difference between the performances of the two classifiers as demonstrated by the results of McNemar’s test (chi-squared; χ² = 0.14, and 0.03 when all the remaining ASIA Eagle wavebands, after removing the noisy ones and the most important wavebands were used, respectively). This study concludes that AISA Eagle data classified using RF and SVM algorithms provide relatively accurate information that is important to the forest industry for making informed decision regarding pine plantations health protocols.     

Dry-season vegetation mass and cover fraction from SWIR1.6 and SWIR2.1 band ratio: Ground-radiometer and MODIS data in the Sahel

The potential of the short-wave infrared (SWIR) bands to detect dry-season vegetation mass and cover fraction is investigated with ground radiometry and MODIS data, confronted to vegetation data collected in rangeland and cropland sites in the Sahel (Senegal, Niger, Mali). The ratio of the 1.6 and 2.1 μm bands (called STI) acquired with a ground radiometer proved well suited for grassland mass estimation up to 2500 kg/ha with a linear relation (r² = 0.89). A curvilinear regression is accurate for masses ranging up to 3500 kg/ha. STI proved also well suited to retrieve vegetation cover fraction in crop fields, fallows and rangelands. Such dry-season monitoring, with either ground or satellite data, has important applications for forage, erosion risk and fire risk assessment in semi-arid areas.     

Retrieval of leaf water content spanning the visible to thermal infrared spectra

The objective of this study was to investigate the entire spectra (from visible to the thermal infrared; 0.390–14.0 μm) to retrieve leaf water content in a consistent manner. Narrow-band spectral indices (calculated from all possible two band combinations) and a partial least square regression (PLSR) were used to assess the strength of each spectral region. The coefficient of determination (R²) and root mean square error (RMSE) were used to report the prediction accuracy of spectral indices and PLSR models. In the visible-near infrared and shortwave infrared (VNIR–SWIR), the most accurate spectral index yielded R² of 0.89 and RMSE of 7.60%, whereas in the mid infrared (MIR) the highest R² was 0.93 and RMSE of 5.97%. Leaf water content was poorly predicted using two-band indices developed from the thermal infrared (R² = 0.33). The most accurate PLSR model resulted from MIR reflectance spectra (R² = 0.96, RMSE = 4.74% and RMSE cross validation RMSE_CV = 6.17%) followed by VNIR–SWIR reflectance spectra (R² = 0.91, RMSE = 6.90% and RMSE_CV = 7.32%). Using thermal infrared (TIR) spectra, the PLSR model yielded a moderate retrieval accuracy (R² = 0.67, RMSE = 13.27% and RMSE_CV = 16.39%). This study demonstrated that the mid infrared (MIR) and shortwave infrared (SWIR) domains were the most sensitive spectral region for the retrieval of leaf water content.     

Comparative analysis of different retrieval methods for mapping grassland leaf area index using airborne imaging spectroscopy

Fine scale maps of vegetation biophysical variables are useful status indicators for monitoring and managing national parks and endangered habitats. Here, we assess in a comparative way four different retrieval methods for estimating leaf area index (LAI) in grassland: two radiative transfer model (RTM) inversion methods (one based on look-up-tables (LUT) and one based on predictive equations) and two statistical modelling methods (one partly, the other entirely based on in situ data). For prediction, spectral data were used that had been acquired over Majella National Park in Italy by the airborne hyperspectral HyMap instrument. To assess the performance of the four investigated models, the normalized root mean squared error (nRMSE) and coefficient of determination (R²) between estimates and in situ LAI measurements are reported (n = 41). Using a jackknife approach, we also quantified the accuracy and robustness of empirical models as a function of the size of the available calibration data set. The results of the study demonstrate that the LUT-based RTM inversion yields higher accuracies for LAI estimation (R² = 0.91, nRMSE = 0.18) as compared to RTM inversions based on predictive equations (R² = 0.79, nRMSE = 0.38). The two statistical methods yield accuracies similar to the LUT method. However, as expected, the accuracy and robustness of the statistical models decrease when the size of the calibration database is reduced to fewer samples. The results of this study are of interest for the remote sensing community developing improved inversion schemes for spaceborne hyperspectral sensors applicable to different vegetation types. The examples provided in this paper may also serve as illustrations for the drawbacks and advantages of physical and empirical models.     

Detection and mapping the spatial distribution of bracken fern weeds using the Landsat 8 OLI new generation sensor

Bracken fern is an invasive plant that presents serious environmental, ecological and economic problems around the world. An understanding of the spatial distribution of bracken fern weeds is therefore essential for providing appropriate management strategies at both local and regional scales. The aim of this study was to assess the utility of the freely available medium resolution Landsat 8 OLI sensor in the detection and mapping of bracken fern at the Cathedral Peak, South Africa. To achieve this objective, the results obtained from Landsat 8 OLI were compared with those derived using the costly, high spatial resolution WorldView-2 imagery. Since previous studies have already successfully mapped bracken fern using high spatial resolution WorldView-2 image, the comparison was done to investigate the magnitude of difference in accuracy between the two sensors in relation to their acquisition costs. To evaluate the performance of Landsat 8 OLI in discriminating bracken fern compared to that of Worldview-2, we tested the utility of (i) spectral bands; (ii) derived vegetation indices as well as (iii) the combination of spectral bands and vegetation indices based on discriminant analysis classification algorithm. After resampling the training and testing data and reclassifying several times (n = 100) based on the combined data sets, the overall accuracies for both Landsat 8 and WorldView-2 were tested for significant differences based on Mann-Whitney U test. The results showed that the integration of the spectral bands and derived vegetation indices yielded the best overall classification accuracy (80.08% and 87.80% for Landsat 8 OLI and WorldView-2 respectively). Additionally, the use of derived vegetation indices as a standalone data set produced the weakest overall accuracy results of 62.14% and 82.11% for both the Landsat 8 OLI and WorldView-2 images. There were significant differences {U (100) = 569.5, z = −10.8242, p < 0.01} between the classification accuracies derived based on Landsat OLI 8 and those derived using WorldView-2 sensor. Although there were significant differences between Landsat and WorldView-2 accuracies, the magnitude of variation (9%) between the two sensors was within an acceptable range. Therefore, the findings of this study demonstrated that the recently launched Landsat 8 OLI multispectral sensor provides valuable information that could aid in the long term continuous monitoring and formulation of effective bracken fern management with acceptable accuracies that are comparable to those obtained from the high resolution WorldView-2 commercial sensor.