Search

Accurate information about phenological stages is essential for canola field management practices such as irrigation, fertilization, and harvesting. Previous studies in canola phenology monitoring focused mainly on the flowering stage, using its apparent structure features and colors. Additional phenological stages have been largely overlooked. The objective of this study was to improve a shape-model method (SMM) for extracting winter canola phenological stages from time-series top-of-canopy reflectance images collected by an unmanned aerial vehicle (UAV). The transformation equation of the SMM was refined to account for the multi-peak features of the temporal dynamics of three vegetation indices (VIs) (NDVI, EVI, and CIred-edge). An experiment with various seeding scenarios was conducted, including four different seeding dates and three seeding densities. Three mathematical functions: asymmetric Gaussian function (AGF), Fourier function, and double logistic function, were employed to fit time-series vegetation indices to extract information about phenological stages. The refined SMM effectively estimated the phenological stages of canola, with a minimum root mean square error (RMSE) of 3.7 days for all phenological stages. The AGF function provided the best fitting performance, as it captured multiple peaks in the growth dynamics characteristics for all seeding date scenarios using four scaling parameters. For the three selected VIs, CIred-edge achieved the greatest accuracy in estimating the phenological stage dates. This study demonstrates the high potential of the refined SMM for estimating winter canola phenology.

The pulse crop lentil (Lens culinaris Medik.) is often grown in crop rotations to provide nitrogen (N) and water benefits for subsequent crops. Lentil yields vary greatly with environmental factors and management. A reliable crop model for lentil could assist efforts to assess the effects of management practices to mitigate environmental stresses and maximize lentil yields. However, few crop models simulate the development and growth of lentil. In this study, we adapted the CSM-CROPGRO model in the Decision Support System for Agrotechnology Transfer to simulate lentil development and growth based on data collected from six experiments conducted from 2001 to 2021 globally. The initial parameter values taken from faba bean (Vicia faba L.) were modified based on reported information and analysis of observed data. Those values were fine-tuned to minimize the gaps between the simulated and observed crop attributes. The model simulated well development stages with root mean square error (RMSE) of 4 days and aboveground biomass with normalized root mean square error (nRMSE ≤ 23%). Seed yields were generally well simulated across experiments in calibration and validation (nRMSE = 19%) datasets, except for overestimation under the humid environment of Quebec in Canada, which may have resulted from excessive vegetative growth. The underlying mechanisms leading to excessive vegetative growth need to be explored further and included in the model for evaluating the adaptability of lentils to specific regions. Overall, the CSM-CROPGRO-Lentil model is ready for simulating lentil production under various scenarios, which may identify ways to improve the productivity and resiliency of cropping systems that include lentil. [ABSTRACT FROM AUTHOR]

With the wide availability of remotely sensed data from various sensors, fusion-based tree species classification approaches have emerged as a prominent and ongoing research topic. However, most recent studies primarily focused on combining multisource data at the feature level, while few systematically examined their positive or negative contributions to tree species classification. This study aimed to investigate fusion approaches at the feature and decision levels deployed with support vector machine and random forest algorithms to classify five dominant tree species: Norway maple, honey locust, Austrian pine, white spruce, and blue spruce in individual crowns. Spectral, textural, and structural features derived from multispectral imagery (MSI), a very high-resolution panchromatic image (PAN), and LiDAR data were systematically exploited to assess their contributions to accurate classifications. Among the various classification schemes that were explored, both feature- and decision-level fusion approaches demonstrated significant improvements in tree species classification compared with the utilization of MSI (0.7), PAN (0.74), or LiDAR (0.8) in isolation. Notably, the decision-level fusion approach achieved the highest overall accuracies (0.86 for SVM and 0.84 for RF) and kappa coefficients (0.82 for SVM and 0.79 for RF). The misclassification analysis of fusion approaches highlighted the potential and flexibility of decision-level fusion in tree species classification.

Height is a key factor in monitoring the growth status and rate of crops. Compared with large-scale satellite remote sensing images and high-cost LiDAR point cloud, the point cloud generated by the Structure from Motion (SfM) algorithm based on UAV images can quickly estimate crop height in the target area at a lower cost. However, crop leaves gradually start to cover the ground from the beginning of the stem elongation stage, making more and more ground points below the canopy disappear in the data. The terrain undulations and outliers will seriously affect the height estimation accuracy. This paper proposed a ground point fitting method to estimate the height of winter wheat based on the UAV SfM point cloud. A canopy slice filter was designed to reduce the interference of middle canopy points and outliers. Random Sample Consensus (RANSAC) was applied to obtain the ground points from the valid filtered point cloud. Then, the missing ground points were fitted according to the known ground points. Furthermore, we achieved crop height monitoring at the stem elongation stage with an R2 of 0.90. The relative root mean squared error (RRMSE) of height estimation was 5.9%, and the relative mean absolute error (RMAE) was 4.6% at the stem elongation stage. This paper proposed the canopy slice filter and fitting missing ground points. It was concluded that the canopy slice filter successfully optimized the extraction of ground points and removed outliers. Fitting the missing ground points simulated the terrain undulations effectively and improved the accuracy.

Soil water deficit and high atmospheric dryness (vapor pressure deficit, VPD) are major environmental limitations on carbon uptake of terrestrial ecosystems. However, it is still unclear how climate seasonality influences seasonal soil water supply and atmospheric water demand, and consequently limits plant photosynthesis. Here, we analyzed the impacts of the seasonal radiation-rainfall coupling on soil moisture limitations versus atmospheric dryness limitations on plant photosynthesis across the Northern Hemisphere north of 15°N, using the eddy covariance data of 83 forest sites and multiple satellite-based data. Our results show that forest photosynthesis is strongly reduced by low soil water availability that is accompanied by a high atmospheric dryness during warm seasons for sites and regions where there is a strong negative covariation between radiation and rainfall availability, which we denote as asynchronous climate. However, under climates with positive covariation between radiation and rainfall availability, i.e. synchronous climate, forest photosynthesis experiences only a small soil water stress, but tends to be limited by high atmospheric dryness during warm seasons. Both the site and regional analyses imply that atmospheric dryness exhibits stronger constraints on forest photosynthesis in synchronous climate over a larger area than in asynchronous climate across the Northern Hemisphere.

Leaf area index (LAI) is a widely used plant biophysical parameter required for modelling plant photosynthesis and crop yield estimation. UAV remote sensing plays an increasingly important role in providing the data source needed for LAI extraction. This study proposed a UAV-derived 3-D point cloud-based method to automatically calculate crop-effective LAI (LAIe). In this method, the 3-D winter wheat point cloud data filtered out of bare ground points was projected onto a hemisphere, and then the gap fraction was calculated through the hemispherical image obtained by projecting the sphere onto a plane. A single-angle inversion method and a multi-angle inversion method were used, respectively, to calculate the LAIe through the gap fraction. The results show a good linear correlation between the calculated LAIe and the field LAIe measured by the digital hemispherical photography method. In particular, the multi-angle inversion method of stereographic projection achieved the highest accuracy, with an R2 of 0.63. The method presented in this paper performs well in LAIe estimation of the main leaf development stages of the winter wheat growth cycle. It offers an effective means for mapping crop LAIe without the need for reference data, which saves time and cost.

Soil moisture is vital for the crop growth and directly affects the crop yield. The conventional synthetic aperture radar (SAR) based soil moisture monitoring is often influenced by vegetation cover and surface roughness. The machine-learning methods are not constrained by physical parameters and have high nonlinear fitting capabilities. In this study, machine-learning methods were applied to estimate soil moisture over winter wheat fields during its growing season. RADARSAT-2 data with quad polarizations and 240 sample plots in the study area were acquired and collected, respectively. In addition to the four linear polarization channels, polarimetric decomposition parameters were extracted to expand the SAR feature space. Three advanced machine-learning models were selected and compared, which were support vector regression, random forests (RF), and gradient boosting regression tree. To improve the performances of the models, three feature-selection methods were compared, which were based on Pearson correlation, support vector machine recursive feature elimination, and RF, respectively. The coefficient of determination (R2) and root-mean-square error (RMSE) were used to compare and assess the performances of those models. The results revealed that polarimetric decomposition parameters were effective in estimating soil moisture, and RF model obtained the highest prediction accuracy (training set: RMSE = 2.44 vol.% and R2 = 0.94; and validation set: RMSE = 4.03 vol.%, and R2 = 0.79). This study finally concluded that using polarimetric decomposition parameters combined with machine-learning and feature-selection methods could effectively estimate soil moisture at a high accuracy, which helps monitor soil moisture across the agricultural field during its growing season.

Soil moisture (Mv) estimation and monitoring over agricultural areas using Synthetic Aperture Radar (SAR) are often affected by vegetation cover during the growing season. Volume scattering and vegetation attenuation can complicate the received SAR backscatter signal when microwave interacts with the vegetation canopy. To address the existing problems, this article employed the model-based polarimetric decomposition method considering the two-way attenuation to remove the volume scattering and vegetation attenuation. A deorientation process of SAR data was applied to remove the influence of randomly distributed target orientation angles before the polarimetric decomposition. To parameterize the two-way attenuation, Radar Vegetation Index derived from the SAR intensity images was adopted. The Dubois model was used to describe backscattering from the underlying bare soil. Since the soil roughness parameters are difficult to measure under vegetation cover, the optimum surface roughness method was used to parameterize the Dubois model. This soil moisture retrieval algorithm was applied to the polarimetric multitemporal RADARSAT-2 SAR data over soybean fields. The validation indicates the root-mean-square error of 9.2 vol.% and 8.2 vol.% at HH and VV polarization, respectively, over the entire soybean growing period, suggesting that the proposed method is capable of reducing the effect of vegetation cover for soil moisture monitoring over the soybean field.

The climatic drivers of leaf phenology and water stress regulation strategies in tropical/subtropical forest biomes are poorly understood on the continental scale. Widespread field observations and remotely sensed plant phenology and physiology data across tropical forest ecosystems at various scales provide new insight into the response of tropical forests to climate change. In this study, we collected seasonal litterfall data from 65 sites across tropical/subtropical America for analysis against remotely sensed photosynthetic and climatic indicators. We proposed an integrated field and remote sensing approach for partitioning three rhythms of forest phenology across tropical and subtropical America in response to various wet- and dry-season variabilities in sunlight, liquid water supply and atmospheric dryness. The results showed that in a dry-season leaf shedding and rejuvenation phenology, trees encounter limited soil water stress, shift to a young canopy by replacing old leaves with new leaves, and regulate water use through the xylem to maximize light capture and productivity. In a wet-season leaf shedding & rejuvenation phenology, trees on the contrary shed and rejuvenate canopy leaves to maximize light use and increase photosynthesis during the sunny wet season. Trees show a dominant stomatal regulation of seasonal water use, resulting in considerable seasonal changes in canopy photosynthesis and transpiration. However, in a dry-season leaf shedding and wet-season rejuvenation phenology, trees shed leaves due to water stresses and grew new leaves to maximize photosynthesis with sufficient water availability. Trees shift from wet-season stomatal regulation to dry-season xylem regulation with fewer canopy leaves. Our findings provide new physiological insight into the mechanism of sunlight availability, liquid water supply and atmospheric dryness in driving leaf phenology and photosynthesis across tropical and subtropical forests.

We evaluated the utility of Terra/MODIS-derived crop metrics for yield estimation across the Canadian Prairies. This study was undertaken at the Census Agriculture Region (CAR) and the Rural Municipality (RM) of the province of Saskatchewan, in three prairie agro-climate zones. We compared MODIS-derived vegetation indices, gross primary productivity (GPP), and net primary productivity (NPP) to the known yields for barley, canola, and spring wheat. Multiple linear regressions were used to assess the relationships between the metrics and yield at the CAR and RM levels for the years 2000 to 2016. Models were evaluated using a leave-one-out cross validation (LOOCV) approach. Results showed that vegetation indices at crop peak growing stages were better predictors of yield than GPP or NPP, and EVI2 was better than NDVI. Using seasonal maximum EVI2, CAR-level crop yields can be estimated with a relative root-mean-square-error (RRMSE) of 14-20% and a Nash-Sutcliffe model efficiency coefficient (NSE) of 0.53-0.70, though the exact relationship varies by crop type and agro-climate zone. LOOCV showed the stability of the models across different years, although interannual fluctuations of estimation accuracy were observed. Assessments using RM-level yields showed slightly reduced accuracy, with NSE of 0.37-0.66, and RRMSE of 18-28%. The best performing models were used to map annual crop yields at the Soil Landscapes of Canada (SLC) polygon level. The results indicated that the models could perform well at both spatial scales, and thus, could be used to disaggregate coarse resolution crop yields to finer spatial resolutions using MODIS data.

Within-field variation of leaf area index (LAI) plays an essential role in field crop monitoring and yield forecasting. Although unmanned aerial vehicle (UAV)-based optical remote sensing method can overcome the spatial and temporal resolution limitations associated with satellite imagery for fine-scale within-field LAI estimation of field crops, image correction and calibration of UAV data are very challenging. In this study, a physical-based method was proposed to automatically calculate crop effective LAI (LAIe) using UAV-based 3-D point cloud data. Regular high spatial resolution RGB images were used to generate point cloud data for the study area. The proposed method, simulated observation of point cloud (SOPC), was designed to obtain the 3-D spatial distribution of vegetation and bare ground points and calculate the gap fraction and LAIe from a UAV-based 3-D point cloud dataset at vertical, 57.5°, and multiview angle of a winter wheat field in London, Ontario, Canada. Results revealed that the derived LAIe using the SOPC multiview angle method correlates well with the LAIe derived from ground digital hemispherical photography, R2 = 0.76. The root mean square error and mean absolute error for the entire experiment period from May 11 to May 27 were 0.19 and 0.14, respectively. The newly proposed method performs well for LAIe estimation during the main leaf development stages (BBCH 20-39) of the growth cycle. This method has the potential to become an alternative approach for crop LAIe estimation without the need for ground-based reference measurements, hence save time and money.

The articles in this special section focus on the use of digital technology in agricultural research and applications. Sustainability of global agricultural and food systems is one of the prominent factors for peaceful future of the world in next few decades. Although agriculture is the main part of the global food supply chain, it is under rising pressure of global climate change, environmental deterioration, and falling per-capita arable land. Efficiency and sustainability management at all levels of agricultural planning and production appear as the most promising balancing factor for the short and medium terms. For this reason, timely and accurate information about the current conditions and future predictions in agriculture and the related resources become more important than ever. We are living in an age that the annual production number of transistors in microprocessors is more than the number of wheat grains produced in the same year. This is an indication of increasing data processing capability and decreasing cost. On the other hand, the number of Internet connected devices is estimated to be more than 20 billion now and is increasing rapidly, and Intent-of-Things (IoT) devices have the highest share in this rising trend. While it may be not able to solve long-term global food sustainability issue, the rapid increase in data collection and processing capabilities for agricultural monitoring and prediction may remediate the issue at least in short and middle terms.

Genetic variation among populations within plant species can have huge impact on canopy biochemistry and structure across broad spatial scales. Since canopy spectral reflectance is determined largely by canopy biochemistry and structure, spectral reflectance can be used as a means to capture the variability of th genetic characteristics of plant species. In this study, we used spectral measurements of Bermuda grass [Cynodon dactylon (L.) Pers.] at both the leaf and canopy levels to characterize the variability of plant traits pertinent to phylogeographic variation along the longitudinal and latitudinal gradients. An integration of airborne multispectral and hyperspectral data allows for the exploitation of spectral variations to discriminate between the five different genotypic groups using ANOVA and RF models. We evaluated the spectral variability among high-latitude genotypic groups and other groups along the latitudinal gradients and assessed spectral variability along longitudinal gradients. Spectral difference was observed between genetic groups from the northern regions and those from other regions along the latitudinal gradient, which indicated the usefulness of spectral signatures for discriminating between genetic groups. The canopy spectral reflectance was better suited to discriminate between genotypes of Bermuda grass across multiple scales than leaf spectral data, as assessed using random forest models. The use of spectral reflectance, derived from remote sensing, for studying genetic variability across landscapes is becoming an emerging research topic, with the potential to monitor and forecast phenology, evolution and biodiversity.

Delineation of agricultural fields is desirable for operational monitoring of agricultural production and is essential to support food security. Due to large within-class variance of pixel values and small inter-class difference, automated field delineation remains to be a challenging task. In this study, a strategy is proposed to effectively address this issue. Firstly, a framework was developed using the Canny operator connected with the Watershed segmentation algorithm (CW) to quickly label the training dataset, which minimizes the workload of dataset generation in comparison with the commonly used manual vectorization. Secondly, a CW-trained deep semantic segmentation network, recurrent residual U-Net, was selected to mine the low level and deep semantic features. Finally, a boundary connecting method (to integrate fragmented boundaries) was used to generate the agricultural field boundary. The proposed methods are tested over smallholder agricultural landscape in Heilongjiang province, China, using Sentinel-2 imagery. Compared with the U-Net (overall accuracy (OA) 82.18%), the residual U-Net (ResU-Net, with OA 85.78%), traditional object-based image analysis (OBIA, with OA about 82%), and the existing 10-m resolution global land cover map (FROM-GLC10), the proposed method shows an improved performance (OA 89.28%, and Kappa 0.85). The successful application of the proposed method suggests that the recurrent residual U-Net model has great universality in agricultural field boundary extraction, and the automated technique has the potential of being applied to other regions.

Timely and accurate mapping crops is essential for agricultural management, policy making and food security. The smallholder agricultural systems lead to large number of fragmented and heterogeneous landscape, making fine crop mapping a huge challenge. Feature and classifier selection are two important influencing factors in crop classification. However, there are few systematic tests to determine the specific features and classifiers needed for heterogeneous agricultural landscapes. In this study, 24 candidate spectral features, 8 spatial texture features from the red-edge (RE) and near-infrared (NIR) bands, and 4 supervised classifiers (i.e. random forest (RF), support vector machine (SVM), artificial neural network (ANN), and maximum likelihood classifier (MLC)) were used for crop mapping. 60 spatially heterogeneous landscapes in Heilongjiang Province, northeastern China were selected as the study areas, evaluated by compositional heterogeneity (homogeneity index, HOI) and configurational heterogeneity (splitting index, SPLIT). The results summarized a look-up table for searching the optimum classification features and classifiers in different landscape, providing a reference for future crop classifications and preventing the consumption of computing time. The results revealed that (1) an optimal feature-subset (with reduced data volume by 65%) can achieve high-accuracy crop mapping in heterogeneous regions. (2) The optimum type and number of features and classifiers are landscape sensitive. When a specific accuracy was required, homogenous regions need a smaller number of features and a simple MLC could meet the requirement. (3) The impact of configurational heterogeneity on textural features is more significant, while compositional heterogeneity performs better on spectral Vegetation Indices (VIs). Findings from this study provide a general guideline for crop mapping in plains or fragmented landscape and areas with single or complex planting structure.

Yellow rust (Puccinia striiformis f. sp. Tritici) and aphid (Sitobion avenae F.) are two major biotic factors threatening winter wheat growth in the main growing region in northern China. The goal of this study was to develop a remote sensing based approach to reliably detect and discriminate yellow rust and aphid infection. The study was conducted in the North China Plain in 2017 based on RapidEye satellite images using three supervised classification algorithms, the maximum-likelihood classifier, the support vector machine, and the random forest. An overall accuracy of above 60% for aphid and above 70% for yellow rust can be achieved using a single image (May 7 or 10, 2017) with any of the three algorithms. With multi-temporal images, the overall accuracies both increased for aphid and yellow rust (above 70% and above 78%). Using the image acquired on 23 May 2017, joint infections by yellow rust and aphid can be detected with satisfaction (>73% overall accuracy), although confusion exists between the two infections. This study demonstrates that winter wheat disease/pest infection can be detected with remote sensing technologies, providing decision support to farmers, insurance companies, and government organizations in the agriculture sector.

Soil moisture content (SMC) is an indispensable basic element for crop growth and development in agricultural production. Obtaining accurate information on SMC in real time over large agricultural areas has important guiding significance for crop yield estimation and production management. In this study, the paper reports on the retrieval of SMC from RADARSAT-2 polarimetric SAR data. The proposed SMC retrieval algorithm includes vegetation correction based on a ratio method and roughness correction based on the optimal roughness method. Three vegetation description parameters (i.e., RVI, LAI, and NDVI) serve as vegetation descriptors in the parameterization of the algorithm. To testify the vegetation correction result of the algorithm, the water cloud model (WCM) was compared with the algorithm. The calibrated integrated equation model (CIEM) was employed to describe the backscattering from the underlying soil. To improve the accuracy of SMC retrieval, the CIEM model was optimized by using the optimal roughness parameter and the normalization method of reference incidence angle. Validation against ground measurements showed a high correlation between the measured and estimated SMC when the NDVI serves as vegetation descriptor (R2 = 0.68, RMSE = 4.15 vol.%, p < 0.01). The overall estimation performance of the proposed SMC retrieval algorithm is better than that of the WCM. It demonstrates that the proposed algorithm has an operational potential to estimate SMC over wheat fields during the growing season.

The application of remote sensing technology in grassland monitoring and management has been ongoing for decades. Compared with traditional ground measurements, remote sensing technology has the overall advantage of convenience, efficiency, and cost effectiveness, especially over large areas. This paper provides a comprehensive review of the latest remote sensing estimation methods for some critical grassland parameters, including above-ground biomass, primary productivity, fractional vegetation cover, and leaf area index. Then, the applications of remote sensing monitoring are also reviewed from the perspective of their use of these parameters and other remote sensing data. In detail, grassland degradation and grassland use monitoring are evaluated. In addition, disaster monitoring and carbon cycle monitoring are also included. Overall, most studies have used empirical models and statistical regression models, while the number of machine learning approaches has an increasing trend. In addition, some specialized methods, such as the light use efficiency approaches for primary productivity and the mixed pixel decomposition methods for vegetation coverage, have been widely used and improved. However, all the above methods have certain limitations. For future work, it is recommended that most applications should adopt the advanced estimation methods rather than simple statistical regression models. In particular, the potential of deep learning in processing high-dimensional data and fitting non-linear relationships should be further explored. Meanwhile, it is also important to explore the potential of some new vegetation indices based on the spectral characteristics of the specific grassland under study. Finally, the fusion of multi-source images should also be considered to address the deficiencies in information and resolution of remote sensing images acquired by a single sensor or satellite.

Crop identification and classification are of great significance to agricultural land use management. The physically constrained general model-based decomposition (PCGMD) has proven to be a promising method in comparison with the typical four-component decomposition methods in scattering mechanism interpretation and identifying vegetation types. However, the robustness of PCGMD requires further investigation from the perspective of final applications. This paper aims to validate the efficiency of the PCGMD method on crop classification for the first time. Seven C-band time-series RADARSAT-2 images were exploited, covering the entire growing season over an agricultural region near London, Ontario, Canada. Firstly, the response and temporal evolution of the four scattering components obtained by PCGMD were analyzed. Then, a forward selection approach was applied to achieve the highest classification accuracy by searching an optimum combination of multi-temporal SAR data with the random forest (RF) algorithm. For comparison, the general model-based decomposition method (GMD), the original and its three improved Yamaguchi four-component decomposition approaches (Y4O, Y4R, S4R, G4U), were used in all tests. The results reveal that the PCGMD method is highly sensitive to seasonal crop changes and matches well with the real physical characteristics of the crops. Among all test methods used, the PCGMD method using six images obtained the optimum classification performance, reaching an overall accuracy of 91.83%.

The authors wish to make the following changes to their paper [...]

Many African countries are facing increasing risks of food insecurity due to rising populations. Accurate and timely information on the spatial distribution of cropland is critical for the effective management of crop production and yield forecast. Most recent cropland products (2015 and 2016) derived from multi-source remote sensing data are available for public use. However, discrepancies exist among these cropland products, and the level of discrepancy is particularly high in several Africa regions. The overall goal of this study was to identify and assess the driving factors contributing to the spatial discrepancies among four cropland products derived from remotely sensed data. A novel approach was proposed to evaluate the spatial agreement of these cropland products and assess the impact of environmental factors such as elevation dispersion, field size, land-cover richness and frequency of cloud cover on these spatial differences. Results from this study show that the overall accuracies of the four cropland products are below 65%. In particular, large disagreements are seen on datasets covering Sahel zone and along the West African coasts. This study has identified land-cover richness as the driving factor with the largest contribution to the spatial disagreement among cropland products over Africa, followed by the high frequency of cloud cover, small and fragmented field size, and elevation complexity. To improve the accuracy of future cropland products for African regions, the data producers are encouraged to take a multi-classification approach and incorporate multi-sensors into their cropland mapping processes.

Agriculture is an important sector in Canada, and annual crop inventories are required in many agricultural applications. Multi-temporal polarimetric synthetic aperture radar (SAR) data have great potential in crop classification due to its less dependency on weather condition. This study, for the first time, investigated the effects of the Minimum Noise Fraction (MNF) transformation of multi-temporal RADARSAT-2 polarimetric SAR data on the performance of cropland classification through the discussing of the performance of different polarimetric SAR parameter sets, and the impact of the timing of RADARSAT-2 datasets in southwestern Ontario. The random forest classifier was adopted due to its excellent ability in crop classification. The results illustrated that the elements of coherency matrix performed the best in agricultural land cover classification. The multi-temporal polarimetric SAR data acquired from the end of June to November gave the best classification accuracy, and an overall accuracy of 90% can be achieved using two images acquired in the middle of September and October. The MNF transformation can further improve the classification accuracy, and this accuracy was competitive with the accuracy produced using the integration of optical and polarimetric SAR data.

In recent years, the delay in sowing has become a major obstacle to high wheat yield in Jiangsu Province, one of the major wheat producing areas in China; hence, it is necessary to screen wheat varieties are resilient for late sowing. This study aimed to provide an effective, fast, and non-destructive monitoring method of soil plant analysis development (SPAD) values, which can represent leaf chlorophyll contents, for late-sown winter wheat variety screening. This study acquired multispectral images using an unmanned aerial vehicle (UAV) at the overwintering stage of winter wheat growth, and further processed these images to extract reflectance of five single spectral bands and calculated 26 spectral vegetation indices. Based on these 31 variables, this study combined three variable selection methods (i.e., recursive feature elimination (RFE), random forest (RF), and Pearson correlation coefficient (r)) with four machine learning algorithms (i.e., random forest regression (RFR), linear kernel-based support vector regression (SVR), radial basis function (RBF) kernel-based SVR, and sigmoid kernel-based SVR), resulted in seven SVR models (i.e., RFE-SVR_linear, RF-SVR_linear, RF-SVR_RBF, RF-SVR_sigmoid, r-SVR_linear, r-SVR_RBF, and r-SVR_sigmoid) and three RFR models (i.e., RFE-RFR, RF-RFR, and r-RFR). The performances of the 10 machine learning models were evaluated and compared with each other according to the achieved coefficient of determination (R2), residual prediction deviation (RPD), root mean square error (RMSE), and relative RMSE (RRMSE) in SPAD estimation. Of the 10 models, the best one was the RF-SVR_sigmoid model, which was the combination of the RF variable selection method and the sigmoid kernel-based SVR algorithm. It achieved high accuracy in estimating SPAD values of the wheat canopy (R2 = 0.754, RPD = 2.017, RMSE = 1.716 and RRMSE = 4.504%). The newly developed UAV- and machine learning-based model provided a promising and real time method to monitor chlorophyll contents at the overwintering stage, which can benefit late-sown winter wheat variety screening.

Assessing the spatial distribution of soil heavy metals in urban areas in relation to land use, lithology and landform may provide insights for soil quality monitoring. This study evaluated the spatial distribution, the sources and the extent of heavy metal(loid)s in the topsoil of Fuzhou city, China. A combination of GIS and multivariate approaches was used to determine the spatial distribution and the sources of heavy metals. Additionally, analysis of variance was used to determine the variability of selected heavy metals across land use, landform, and lithology. The result show that the mean concentrations of Cd, Zn, As and Pb were higher than background values. Most of the heavy metals had significant correlations with each other. In particular, V and Fe (0.84 at p < 0.01) and Ni and Cr (0.74 at p < 0.01) had strong correlations, while Cu and Fe (0.68 at p < 0.01), Cu and V (0.63 at p < 0.01), Cu and Co (0.52 at p < 0.01), Zn and Ni (0.51 at p < 0.01), Co and Fe (0.54 at p < 0.01), and Cu and Zn (0.55 at p < 0.01) had moderate correlations. Arsenic, Cu, and Zn had significant positive correlations with total nitrogen (TN). Similarly, arsenic, Zn and Cr had positive correlations with total carbon (TC), while Co had negative correlations with TN and TC at p < 0.01. The peak values for Cr, Ni, Pb, Mn, and Zn were observed in the intensively urbanized central and eastern parts of the study area, suggesting that the main sources might be anthropogenic activities. Agricultural land use had the highest content of Cd, which may be attributed to the historical long-term application of agrochemicals in the area. Additionally, its content was significantly higher in agricultural land use with shale lithology, implying that shale lithology was a key geogenic source for Cd of soils in the study area. Pb content was affected by urban land use, which may be attributed to intensive human activities such as emissions from vehicles, industrial effluents, mining activities, and other discharges. The results show the high spatial variability of heavy metal(loid)s, implying that the soils in the study area were highly influenced by both geogenic variability and human activities. Moreover, land use and lithology had significant impacts on the variability of Cd, As and Pb. Sustainable agricultural practices and urban management are recommended to sustain the eco-environment of coastal city.

Rice false smut (RFS), caused by Ustilaginoidea virens, is a significant grain disease in rice that can lead to reduced yield and quality. In order to obtain spatiotemporal change information, multitemporal hyperspectral UAV data were used in this study to determine the sensitive wavebands for RFS identification, 665–685 and 705–880 nm. Then, two methods were used for the extraction of rice false smut-infected areas, one based on spectral similarity analysis and one based on spectral and temporal characteristics. The final overall accuracy of the two methods was 74.23 and 85.19%, respectively, showing that the second method had better prediction accuracy. In addition, the classification results of the two methods show that the areas of rice false smut infection had an expanding trend over time, which is consistent with the natural development law of rice false smut, and also shows the scientific nature of the two methods.

In North Korea, reliable and timely information on crop acreage and spatial distribution is hard to obtain. In this study, we developed a fast and robust method to estimate crop acreage in North Korea using time-series normalized difference vegetation index (NDVI) derived from the Moderate Resolution Imaging Spectroradiometer (MODIS) data. We proposed a method to identify crop type based on NDVI phenology features using data collected in other areas with similar agri-environmental conditions to mitigate the shortage of ground truth data. Eventually the classification map (MODIScrop) was assessed using the Food and Agriculture Organization (FAO) statistical data and high-resolution crop classification maps derived from one Landsat scene (LScrop). The Pareto boundary method was used to assess the accuracy and crop distribution of the MODIScrop maps. Results showed that acreage derived from the MODIScrop maps was generally consistent with that reported in the FAO data (a relative error

The well-documented energy balance dynamics within forest ecosystems are poorly implemented in studies of the biophysical effects of forests. This results in limitations to the accurate quantification of forest cooling/warming on local air temperature. Taking into consideration the forest air space, this study proposes a three-layered (canopy, forest air space and soil [CAS]) land surface energy balance model to simulate air temperature within forest spaces (Taf) and subsequently to evaluate its biophysical effects on forest cooling/warming, i.e., the air temperature gradient (∆Ta) between the Taf and air temperature of open spaces (Tao) (∆Ta = Taf − Tao). We test the model using field data for 23 sites across 10 cities worldwide; the model shows satisfactory performance with the test data. High-latitude forests show greater seasonal dynamics of ∆Ta, generating considerable cooling of local air temperatures in warm seasons but minimal cooling or even warming effects during cool seasons, while low-latitude tropical forests always exert cooling effects with less interannual variability. The interannual dynamics of ∆Ta are significantly related to the seasonality of solar geometry and canopy leaf phenology. The differences between forest canopy temperature (Tc) and Tao, which are the two most important terms attributed by the CAS model in impacting Taf, explain a large part of forest cooling and warming (May–July: R2 = 0.35; November–January: R2 = 0.51). The novel CAS model provides a feasible way to represent the energy balance within forest ecosystems and to assess its impacts on local air temperatures globally. Keywords: Forest cooling/warming, Forest biophysical effects, Radiative transfer equation, Land-atmosphere model, Local environmental health, Climate warming mitigation

Assessing the spatial dynamics of soil organic carbon (SOC) is essential for carbon monitoring. Since variability of SOC is mainly attributed to biophysical land surface variables, integrating a compressive set of such indices may support the pursuit of an optimum set of predictor variables. Therefore, this study was aimed at predicting the spatial distribution of SOC in relation to remotely sensed variables and other covariates. Hence, the land surface variables were combined from remote sensing, topographic, and soil spectral sources. Moreover, the most influential variables for prediction were selected using the random forest (RF) and classification and regression tree (CART). The results indicated that the RF model has good prediction performance with corresponding R2 and root-mean-square error (RMSE) values of 0.96 and 0.91 mg·g−1, respectively. The distribution of SOC content showed variability across landforms (CV = 78.67%), land use (CV = 93%), and lithology (CV = 64.67%). Forestland had the highest SOC (13.60 mg·g−1) followed by agriculture (10.43 mg·g−1), urban (9.74 mg·g−1), and water body (4.55 mg·g−1) land uses. Furthermore, soils developed in bauxite and laterite lithology had the highest SOC content (14.69 mg·g−1). The SOC content was remarkably lower in soils developed in sandstones; however, the values obtained in soils from the rest of the lithologies could not be significantly differentiated. The mean SOC concentration was 11.70 mg·g−1, where the majority of soils in the study area were classified as highly humus and extremely humus. The soils with the highest SOC content (extremely humus) were distributed in the mountainous regions of the study area. The biophysical land surface indices, brightness removed vegetation indices, topographic indices, and soil spectral bands were the most influential predictors of SOC in the study area. The spatial variability of SOC may be influenced by landform, land use, and lithology of the study area. Remotely sensed predictors including land moisture, land surface temperature, and built-up indices added valuable information for the prediction of SOC. Hence, the land surface indices may provide new insights into SOC modeling in complex landscapes of warm subtropical urban regions.

Multitemporal polarimetric synthetic aperture radar (PolSAR) has proven as a very effective technique in agricultural monitoring and crop classification. This study presents a comprehensive evaluation of crop monitoring and classification over an agricultural area in southwestern Ontario, Canada. The time-series RADARSAT-2 C-Band PolSAR images throughout the entire growing season were exploited. A set of 27 representative polarimetric observables categorized into ten groups was selected and analyzed in this research. First, responses and temporal evolutions of each of the polarimetric observables over different crop types were quantitatively analyzed. The results reveal that the backscattering coefficients in cross-pol and Pauli second channel, the backscattering ratio between HV and VV channels (HV/VV), the polarimetric decomposition outputs, the correlation coefficient between HH and VV channelρ ρHHVV, and the radar vegetation index (RVI) show the highest sensitivity to crop growth. Then, the capability of PolSAR time-series data of the same beam mode was also explored for crop classification using the Random Forest (RF) algorithm. The results using single groups of polarimetric observables show that polarimetric decompositions, backscattering coefficients in Pauli and linear polarimetric channels, and correlation coefficients produced the best classification accuracies, with overall accuracies (OAs) higher than 87%. A forward selection procedure to pursue optimal classification accuracy was expanded to different perspectives, enabling an optimal combination of polarimetric observables and/or multitemporal SAR images. The results of optimal classifications show that a few polarimetric observables or a few images on certain critical dates may produce better accuracies than the whole dataset. The best result was achieved using an optimal combination of eight groups of polarimetric observables and six SAR images, with an OA of 94.04%. This suggests that an optimal combination considering both perspectives may be valuable for crop classification, which could serve as a guideline and is transferable for future research.

We evaluate the potential of using a process-based ecosystem model (BEPS) for crop biomass mapping at 20 m resolution over the research site in Manitoba, western Canada driven by spatially explicit leaf area index (LAI) retrieved from Sentinel-2 spectral reflectance throughout the entire growing season. We find that overall, the BEPS-simulated crop gross primary production (GPP), net primary production (NPP), and LAI time-series can explain 82%, 83%, and 85%, respectively, of the variation in the above-ground biomass (AGB) for six selected annual crops, while an application of individual crop LAI explains only 50% of the variation in AGB. The linear relationships between the AGB and these three indicators (GPP, NPP and LAI time-series) are rather high for the six crops, while the slopes of the regression models vary for individual crop type, indicating the need for calibration of key photosynthetic parameters and carbon allocation coefficients. This study demonstrates that accumulated GPP and NPP derived from an ecosystem model, driven by Sentinel-2 LAI data and abiotic data, can be effectively used for crop AGB mapping; the temporal information from LAI is also effective in AGB mapping for some crop types.

This study presents a demonstration of the applicability of machine learning techniques for the retrieval of crop height in corn fields using space-borne PolSAR (Polarimetric Synthetic Aperture Radar) data. Multi-year RADARSAT-2 C-band data acquired over agricultural areas in Canada, covering the whole corn growing period, are exploited. Two popular machine learning regression methods, i.e., Random Forest Regression (RFR) and Support Vector Regression (SVR) are adopted and evaluated. A set of 27 representative polarimetric parameters are extracted from the PolSAR data and used as input features in the regression models for height estimation. Furthermore, based on the unique capability of the RFR method to determine variable importance contributing to the regression, a smaller number of polarimetric features (6 out of 27 in our study) are selected in the final regression models. Results of our study demonstrate that PolSAR observables can produce corn height estimates with root mean square error (RMSE) around 40–50 cm throughout the growth cycle. The RFR approach shows better overall accuracy in corn height estimation than the SVR method in all tests. The six selected polarimetric features by variable importance ranking can generate better results. This study provides a new perspective on the use of PolSAR data in retrieving agricultural crop height from space.

Accurate and continuous monitoring of leaf area index (LAI), a widely-used vegetation structural parameter, is crucial to characterize crop growth conditions and forecast crop yield. Meanwhile, advancements in collecting field LAI measurements have provided strong support for validating remote-sensing-derived LAI. This paper evaluates the performance of LAI retrieval from multi-source, remotely sensed data through comparisons with continuous field LAI measurements. Firstly, field LAI was measured continuously over periods of time in 2018 and 2019 using LAINet, a continuous LAI measurement system deployed using wireless sensor network (WSN) technology, over an agricultural region located at the Heihe watershed at northwestern China. Then, cloud-free images from optical satellite sensors, including Landsat 7 the Enhanced Thematic Mapper Plus (ETM+), Landsat 8 the Operational Land Imager (OLI), and Sentinel-2A/B Multispectral Instrument (MSI), were collected to derive LAI through inversion of the PROSAIL radiation transfer model using a look-up-table (LUT) approach. Finally, field LAI data were used to validate the multi-temporal LAI retrieved from remote-sensing data acquired by different satellite sensors. The results indicate that good accuracy was obtained using different inversion strategies for each sensor, while Green Chlorophyll Index (CIgreen) and a combination of three red-edge bands perform better for Landsat 7/8 and Sentinel-2 LAI inversion, respectively. Furthermore, the estimated LAI has good consistency with in situ measurements at vegetative stage (coefficient of determination R2 = 0.74, and root mean square error RMSE = 0.53 m2 m−2). At the reproductive stage, a significant underestimation was found (R2 = 0.41, and 0.89 m2 m−2 in terms of RMSE). This study suggests that time-series LAI can be retrieved from multi-source satellite data through model inversion, and the LAINet instrument could be used as a low-cost tool to provide continuous field LAI measurements to support LAI retrieval.

Green leaf area index (LAI) is an important variable related to crop growth. Accurate and timely information on LAI is essential for developing suitable field management strategies to mitigate risk and boost yield. Several remote sensing (RS) based methods have been recently developed to estimate LAI at the regional scale. However, the performance of these methods tends to be affected by the quality of RS data, especially when time-series LAI are required. For crop LAI estimation, supplementary growth information from crop model is helpful to address this issue. In this study, we focus on the regional-scale LAI estimations of spring maize for the entire growth season. Using time-series multispectral RS data acquired by an unmanned aerial vehicle (UAV) and the World Food Studies (WOFOST) crop model, three methods were applied at different crop growth stages: empirical method using vegetation index (VI), data assimilation method and hybrid method. The VI-based method and assimilation method were used to generate time-series LAI estimations for the whole crop growth season. Then, a hybrid method specially for the late-stage LAI retrieval was developed by integrating WOFOST model and data assimilation. Using field-collected LAI data in Hongxing Farm in 2014, the performances of these three methods were evaluated. At the early stage, the VI-based method (R2 = 0.63, RMSE = 0.16, n = 36) achieved higher accuracy than the assimilation method (R2 = 0.54, RMSE = 0.52, n = 36), whereas at the mid stage, the assimilation method (R2 = 0.63, RMSE = 0.46, n = 28) showed higher accuracy than the VI-based method (R2 = 0.41, RMSE = 0.51, n = 28). At the late stage, the hybrid method yielded the highest accuracy (R2 = 0.63, RMSE = 0.46, n = 29), compared with the VI-based method (R2 = 0.19, RMSE = 0.43, n = 28) and the assimilation method (R2 = 0.20, RMSE = 0.44, n = 29). Based on the results above, we considered a combination of the three methods, i.e., the VI-based method for the early stage, the assimilation method for the mid stage, and the hybrid method for the late stage, as an ideal strategy for spring-maize LAI estimation for the entire growth season of 2014 in Hongxing Farm, and the accuracy of the combined method over the whole growth season is higher than that of any single method.

Chlorophyll is an essential pigment for photosynthesis in crops, and leaf chlorophyll content can be used as an indicator for crop growth status and help guide nitrogen fertilizer applications. Estimating crop chlorophyll content plays an important role in precision agriculture. In this study, a variable, rate of change in reflectance between wavelengths ‘a’ and ‘b’ (RCRWa-b), derived from in situ hyperspectral remote sensing data combined with four advanced machine learning techniques, Gaussian process regression (GPR), random forest regression (RFR), support vector regression (SVR), and gradient boosting regression tree (GBRT), were used to estimate the chlorophyll content (measured by a portable soil–plant analysis development meter) of rice. The performances of the four machine learning models were assessed and compared using root mean square error (RMSE), mean absolute error (MAE), and coefficient of determination (R2). The results revealed that four features of RCRWa-b, RCRW551.0–565.6, RCRW739.5–743.5, RCRW684.4–687.1 and RCRW667.9–672.0, were effective in estimating the chlorophyll content of rice, and the RFR model generated the highest prediction accuracy (training set: RMSE = 1.54, MAE =1.23 and R2 = 0.95; validation set: RMSE = 2.64, MAE = 1.99 and R2 = 0.80). The GPR model was found to have the strongest generalization (training set: RMSE = 2.83, MAE = 2.16 and R2 = 0.77; validation set: RMSE = 2.97, MAE = 2.30 and R2 = 0.76). We conclude that RCRWa-b is a useful variable to estimate chlorophyll content of rice, and RFR and GPR are powerful machine learning algorithms for estimating the chlorophyll content of rice.

Remote sensing is a useful tool for monitoring spatio-temporal variations of crop morphological and physiological status and supporting practices in precision farming. In comparison with multispectral imaging, hyperspectral imaging is a more advanced technique that is capable of acquiring a detailed spectral response of target features. Due to limited accessibility outside of the scientific community, hyperspectral images have not been widely used in precision agriculture. In recent years, different mini-sized and low-cost airborne hyperspectral sensors (e.g., Headwall Micro-Hyperspec, Cubert UHD 185-Firefly) have been developed, and advanced spaceborne hyperspectral sensors have also been or will be launched (e.g., PRISMA, DESIS, EnMAP, HyspIRI). Hyperspectral imaging is becoming more widely available to agricultural applications. Meanwhile, the acquisition, processing, and analysis of hyperspectral imagery still remain a challenging research topic (e.g., large data volume, high data dimensionality, and complex information analysis). It is hence beneficial to conduct a thorough and in-depth review of the hyperspectral imaging technology (e.g., different platforms and sensors), methods available for processing and analyzing hyperspectral information, and recent advances of hyperspectral imaging in agricultural applications. Publications over the past 30 years in hyperspectral imaging technology and applications in agriculture were thus reviewed. The imaging platforms and sensors, together with analytic methods used in the literature, were discussed. Performances of hyperspectral imaging for different applications (e.g., crop biophysical and biochemical properties’ mapping, soil characteristics, and crop classification) were also evaluated. This review is intended to assist agricultural researchers and practitioners to better understand the strengths and limitations of hyperspectral imaging to agricultural applications and promote the adoption of this valuable technology. Recommendations for future hyperspectral imaging research for precision agriculture are also presented.

Knowledge of sub-field yield potential is critical for guiding precision farming. The recently developed simulated observation of point cloud (SOPC) method can generate high spatial resolution winter wheat effective leaf area index (SOPC-LAIe) maps from the unmanned aerial vehicle (UAV)-based point cloud data without ground-based measurements. In this study, the SOPC-LAIe maps, for the first time, were applied to the simple algorithm for yield estimation (SAFY) to generate the sub-field biomass and yield maps. First, the dry aboveground biomass (DAM) measurements were used to determine the crop cultivar-specific parameters and simulated green leaf area index (LAI) in the SAFY model. Then, the SOPC-LAIe maps were converted to green LAI using a normalization approach. Finally, the multiple SOPC-LAIe maps were applied to the SAFY model to generate the final DAM and yield maps. The root mean square error (RMSE) between the estimated and measured yield is 88 g/m2, and the relative root mean squire error (RRMSE) is 15.2%. The pixel-based DAM and yield map generated in this study revealed clearly the within-field yield variation. This framework using the UAV-based SOPC-LAIe maps and SAFY model could be a simple and low-cost alternative for final yield estimation at the sub-field scale.

Synthetic aperture radar (SAR) is more sensitive to the dielectric properties and structure of the targets and less affected by weather conditions than optical sensors, making it more capable of detecting changes induced by management practices in agricultural fields. In this study, the capability of C-band SAR data for detecting crop seeding and harvest events was explored. The study was conducted for the 2019 growing season in Temiskaming Shores, an agricultural area in Northern Ontario, Canada. Time-series SAR data acquired by Sentinel-1 constellation with the interferometric wide (IW) mode with dual polarizations in VV (vertical transmit and vertical receive) and VH (vertical transmit and horizontal receive) were obtained. interferometric SAR (InSAR) processing was conducted to derive coherence between each pair of SAR images acquired consecutively in time throughout the year. Crop seeding and harvest dates were determined by analyzing the time-series InSAR coherence and SAR backscattering. Variation of SAR backscattering coefficients, particularly the VH polarization, revealed seasonal crop growth patterns. The change in InSAR coherence can be linked to change of surface structure induced by seeding or harvest operations. Using a set of physically based rules, a simple algorithm was developed to determine crop seeding and harvest dates, with an accuracy of 85% (n = 67) for seeding-date identification and 56% (n = 77) for harvest-date identification. The extra challenge in harvest detection could be attributed to the impacts of weather conditions, such as rain and its effects on soil moisture and crop dielectric properties during the harvest season. Other factors such as post-harvest residue removal and field ploughing could also complicate the identification of harvest event. Overall, given its mechanism to acquire images with InSAR capability at 12-day revisiting cycle with a single satellite for most part of the Earth, the Sentinel-1 constellation provides a great data source for detecting crop field management activities through coherent or incoherent change detection techniques. It is anticipated that this method could perform even better at a shorter six-day revisiting cycle with both satellites for Sentinel-1. With the successful launch (2019) of the Canadian RADARSAT Constellation Mission (RCM) with its tri-satellite system and four polarizations, we are likely to see improved system reliability and monitoring efficiency.

Annual crop inventory information is important for many agriculture applications and government statistics. The synergistic use of multi-temporal polarimetric synthetic aperture radar (SAR) and available multispectral remote sensing data can reduce the temporal gaps and provide the spectral and polarimetric information of the crops, which is effective for crop classification in areas with frequent cloud interference. The main objectives of this study are to develop a deep learning model to map agricultural areas using multi-temporal full polarimetric SAR and multi-spectral remote sensing data, and to evaluate the influence of different input features on the performance of deep learning methods in crop classification. In this study, a one-dimensional convolutional neural network (Conv1D) was proposed and tested on multi-temporal RADARSAT-2 and VENµS data for crop classification. Compared with the Multi-Layer Perceptron (MLP), Recurrent Neural Network (RNN) and non-deep learning methods including XGBoost, Random Forest (RF), and Support Vector Machina (SVM), the Conv1D performed the best when the multi-temporal RADARSAT-2 data (Pauli decomposition or coherency matrix) and VENµS multispectral data were fused by the Minimum Noise Fraction (MNF) transformation. The Pauli decomposition and coherency matrix gave similar overall accuracy (OA) for Conv1D when fused with the VENµS data by the MNF transformation (OA = 96.65 ± 1.03% and 96.72 ± 0.77%). The MNF transformation improved the OA and F-score for most classes when Conv1D was used. The results reveal that the coherency matrix has a great potential in crop classification and the MNF transformation of multi-temporal RADARSAT-2 and VENµS data can enhance the performance of Conv1D.