Rice (Oryza sativa L.) is a staple food crop for more than half of the world’s population. Rice production is facing a myriad of problems, including water shortage, climate, and land-use change. Accurate maps of rice growth stages are critical for monitoring rice production and assessing its impacts on national and global food security. Rice growth stages are typically monitored by coarse-resolution satellite imagery. However, it is difficult to accurately map due to the occurrence of mixed pixels in fragmented and patchy rice fields, as well as cloud cover, particularly in tropical countries. To solve these problems, we developed an automated mapping workflow to produce near real-time multi-temporal maps of rice growth stages at a 10-m spatial resolution using multisource remote sensing data (Sentinel-2, MOD13Q1, and Sentinel-1). This study was investigated between 1 June and 29 September 2018 in two (wet and dry) areas of Java Island in Indonesia. First, we built prediction models based on Sentinel-2, and fusion of MOD13Q1/Sentinel-1 using the ground truth information. Second, we applied the prediction models on all images in area and time and separation between the non-rice planting class and rice planting class over the cropping pattern. Moreover, the model’s consistency on the multitemporal map with a 5–30-day lag was investigated. The result indicates that the Sentinel-2 based model classification gives a high overall accuracy of 90.6% and the fusion model MOD13Q1/Sentinel-1 shows 78.3%. The performance of multitemporal maps was consistent between time lags with an accuracy of 83.27–90.39% for Sentinel-2 and 84.15% for the integration of Sentinel-2/MOD13Q1/Sentinel-1. The results from this study show that it is possible to integrate multisource remote sensing for regular monitoring of rice phenology, thereby generating spatial information to support local-, national-, and regional-scale food security applications.
Grasslands are important for their ecological values and for agricultural activities such as livestock production worldwide. Efficient grassland management is vital to these values and activities, and remote sensing technologies are increasingly being used to characterize the spatiotemporal variation of grasslands to support those management practices. For this study, Sentinel-2 satellite imagery was used as an input to develop an open-source and automated monitoring system (Sen2Grass) to gain field-specific grassland information on the national and regional level for any given time range as of January 2016. This system was implemented in a cloud-computing platform (StellaSpark Nexus) designed to process large geospatial data streams from a variety of sources and was tested for a number of parcels from the Haus Riswick experimental farm in Germany. Despite outliers due to fluctuating weather conditions, vegetation index time series suggested four distinct growing cycles per growing season. Established relationships between vegetation indices and grassland yield showed poor to moderate positive trends, implying that vegetation indices could be a potential predictor for grassland biomass and chlorophyll content. However, the inclusion of larger and additional datasets such as Sentinel-1 imagery could be beneficial to developing more robust prediction models and for automatic detection of mowing events for grasslands.
Hyperspectral and multispectral imagery have been demonstrated to have a considerable potential for near real-time monitoring and mapping of grass quality indicators. The objective of this study was to evaluate the efficiency of remote sensing techniques for quantification of aboveground grass biomass (BM) and crude protein (CP) in a temperate European climate such as Ireland. The experiment was conducted on 64 plots and 53 paddocks with varying quantities of nitrogen applied. Hyperspectral imagery (HSI) and multispectral imagery (MSI) were analyzed to develop the prediction models. The MSI data used in this study were captured using an unmanned aircraft vehicle (UAV) and the satellite Sentinel-2, while the HSI data were obtained using a handheld hyperspectral camera. The prediction models were developed using partial least squares regression (PLSR) and stepwise multi-linear regression (MLR). Eventually, the spatial distribution of grass biomass over plots and paddocks was mapped to assess the within-field variability of grass quality metrics. An excellent accuracy was achieved for the prediction of BM and CP using HSI (RPD > 2.5 and R2 > 0.8), and a good accuracy was obtained via MSI-UAV (2 < RPD < 2.5 and R2 > 0.7) for the grass quality indicators. The accuracy of the models calculated using MSI-Sentinel-2 was reasonable for BM prediction and insufficient for CP estimation. The red-edge range of the wavelengths showed the maximum impact on the predictability of grass BM, and the NIR range had the greatest influence on the estimation of grass CP. Both the PLSR and MLR techniques were found to be sufficiently robust for spectral modelling of aboveground BM and CP. The PLSR yielded a slightly better model than MLR. This study suggested that remote sensing techniques can be used as a rapid and reliable approach for near real-time quantitative assessment of fresh grass quality under a temperate European climate.
Soil organic carbon (SOC) loss is one of the main causes of soil degradation in croplands. Thus, spatial and temporal monitoring of SOC is extremely important, both from the environmental and economic perspective. In this regard, the high temporal, spatial, and spectral resolution of the Sentinel-2 data can be exploited for monitoring SOC contents in the topsoil of croplands. In this study, we aim to test the effect of the threshold for a spectral index linked to soil moisture and crop residues on the performance of SOC prediction models using the Multi-Spectral Instrument (MSI) Sentinel-2 and the European Land Use/cover Area frame Statistical survey (LUCAS) topsoil database. The LUCAS spectral data resampled according to MSI/Sentinel-2 bands, which were used to build SOC prediction models combining pairs of the bands. The SOC models were applied to a Sentinel-2 image acquired in North-Eastern Germany after removing the pixels characterized by clouds and green vegetation. Then, we tested different thresholds of the Normalized Burn Ratio 2 (NBR2) index in order to mask moist soil pixels and those with dry vegetation and crop residues. The model accuracy was tested on an independent validation database and the best ratio of performance to deviation (RPD) was obtained using the average between bands B6 and B5 (Red-Edge Carbon Index: RE-CI) (RPD: 4.4) and between B4 and B5 (Red-Red-Edge Carbon Index: RRE-CI) (RPD: 2.9) for a very low NBR2 threshold (0.05). Employing a higher NBR2 tolerance (higher NBR2 values), the mapped area increases to the detriment of the validation accuracy. The proposed approach allowed us to accurately map SOC over a large area exploiting the LUCAS spectral library and, thus, avoid a new ad hoc field campaign. Moreover, the threshold for selecting the bare soil pixels can be tuned, according to the goal of the survey. The quality of the SOC map for each tolerance level can be judged based on the figures of merit of the model.
Soil organic carbon (SOC) is a variable of vital environmental significance in terms of soil quality and function, global food security, and climate change mitigation. Estimation of its content and prediction accuracy on a broader scale remain crucial. Although, spectroscopy under proximal sensing remains one of the best approaches to accurately predict SOC, however, spectroscopy limitation to estimate SOC on a larger spatial scale remains a concern. Therefore, for an efficient quantification of SOC content, faster and less costly techniques are needed, recent studies have suggested the use of remote sensing approaches. The primary aim of this research was to evaluate and compare the capabilities of small Unmanned Aircraft Systems (UAS) for monitoring and estimation of SOC with those obtained from spaceborne (Sentinel-2) and proximal soil sensing (field spectroscopy measurements) on an agricultural field low in SOC content. Nine calculated spectral indices were added to the remote sensing approaches (UAS and Sentinel-2) to enhance their predictive accuracy. Modeling was carried out using various bands/wavelength (UAS (6), Sentinel-2 (9)) and the calculated spectral indices were used as independent variables to generate soil prediction models using five-fold cross-validation built using random forest (RF) and support vector machine regression (SVMR). The correlation regarding SOC and the selected indices and bands/wavelengths was determined prior to the prediction. Our results revealed that the selected spectral indices slightly influenced the output of UAS compared to Sentinel-2 dataset as the latter had only one index correlated with SOC. For prediction, the models built on UAS data had a better accuracy with RF than the two other data used. However, using SVMR, the field spectral prediction models achieved a better overall result for the entire study (log(1/R), RPD = 1.40; R2CV = 0.48; RPIQ = 1.65; RMSEPCV = 0.24), followed by UAS and then Sentinel-2, respectively. This study has shown that UAS imagery can be exploited efficiently using spectral indices.
Accurate measurement of canopy chlorophyll content (CCC) is essential for the understanding of terrestrial ecosystem dynamics through monitoring and evaluating properties such as carbon and water flux, productivity, light use efficiency as well as nutritional and environmental stresses. Information on the amount and distribution of CCC helps to assess and report biodiversity indicators related to ecosystem processes and functional aspects. Therefore, measuring CCC continuously and globally from earth observation data is critical to monitor the status of the biosphere. However, generic and robust methods for regional and global mapping of CCC are not well defined. This study aimed at examining the spatiotemporal consistency and scalability of selected methods for CCC mapping across biomes. Four methods (i.e., radiative transfer models (RTMs) inversion using a look-up table (LUT), the biophysical processor approach integrated into the Sentinel application platform (SNAP toolbox), simple ratio vegetation index (SRVI), and partial least square regression (PLSR)) were evaluated. Similarities and differences among CCC products generated by applying the four methods on actual Sentinel-2 data in four biomes (temperate forest, tropical forest, wetland, and Arctic tundra) were examined by computing statistical measures and spatiotemporal consistency pairwise comparisons. Pairwise comparison of CCC predictions by the selected methods demonstrated strong agreement. The highest correlation (R2 = 0.93, RMSE = 0.4371 g/m2) was obtained between CCC predictions of PROSAIL inversion by LUT and SNAP toolbox approach in a wetland when a single Sentinel-2 image was used. However, when time-series data were used, it was PROSAIL inversion against SRVI (R2 = 0.88, RMSE = 0.19) that showed greatest similarity to the single date predictions (R2 = 0.83, RMSE = 0.17 g/m2) in this biome. Generally, the CCC products obtained using the SNAP toolbox approach resulted in a systematic over/under-estimation of CCC. RTMs inversion by LUT (INFORM and PROSAIL) resulted in a non-biased, spatiotemporally consistent prediction of CCC with a range closer to expectations. Therefore, the RTM inversion using LUT approaches particularly, INFORM for ‘forest’ and PROSAIL for ‘short vegetation’ ecosystems, are recommended for CCC mapping from Sentinel-2 data for worldwide mapping of CCC. Additional validation of the two RTMs with field data of CCC across biomes is required in the future.
Since the onset of agriculture, soils have lost their organic carbon to such an extent that the soil functions of many croplands are threatened. Hence, there is a strong demand for mapping and monitoring critical soil properties and in particular soil organic carbon (SOC). Pilot studies have demonstrated the potential for remote sensing techniques for SOC mapping in croplands. It has, however, been shown that the assessment of SOC may be hampered by the condition of the soil surface. While growing vegetation can be readily detected by means of the well-known Normalized Difference Vegetation Index (NDVI), the distinction between bare soil and crop residues is expressed in the shortwave infrared region (SWIR), which is only covered by two broad bands in Landsat or Sentinel-2 imagery. Here we tested the effect of thresholds for the Cellulose Absorption Index (CAI), on the performance of SOC prediction models for cropland soils. Airborne Prism Experiment (APEX) hyperspectral images covering an area of 240 km2 in the Belgian Loam Belt were used together with a local soil dataset. We used the partial least square regression (PLSR) model to estimate the SOC content based on 104 georeferenced calibration samples (NDVI < 0.26), firstly without setting a CAI threshold, and obtained a satisfactory result (coefficient of determination (R2) = 0.49, Ratio of Performance to Deviation (RPD) = 1.4 and Root Mean Square Error (RMSE) = 2.13 g kgC−1 for cross-validation). However, a cross comparison of the estimated SOC values to grid-based measurements of SOC content within three fields revealed a systematic overestimation for fields with high residue cover. We then tested different CAI thresholds in order to mask pixels with high residue cover. The best model was obtained for a CAI threshold of 0.75 (R2 = 0.59, RPD = 1.5 and RMSE = 1.75 g kgC−1 for cross-validation). These results reveal that the purity of the pixels needs to be assessed aforehand in order to produce reliable SOC maps. The Normalized Burn Ratio (NBR2) index based on the SWIR bands of the MSI Sentinel 2 sensor extracted from images collected nine days before the APEX flight campaign correlates well with the CAI index of the APEX imagery. However, the NBR2 index calculated from Sentinel 2 images under moist conditions is poorly correlated with residue cover. This can be explained by the sensitivity of the NBR2 index to both soil moisture and residues.
Given the continuous increase in the global population, the food manufacturers are advocated to either intensify the use of cropland or expand the farmland, making land cover and land usage dynamics mapping vital in the area of remote sensing. In this regard, identifying and classifying a high-resolution satellite imagery scene is a prime challenge. Several approaches have been proposed either by using static rule-based thresholds (with limitation of diversity) or neural network (with data-dependent limitations). This paper adopts the inductive approach to learning from surface reflectances. A manually labeled Sentinel-2 dataset was used to build a Machine Learning (ML) model for scene classification, distinguishing six classes (Water, Shadow, Cirrus, Cloud, Snow, and Other). This models was accessed and further compared to the European Space Agency (ESA) Sen2Cor package. The proposed ML model presents a Micro-F1 value of 0.84, a considerable improvement when compared to the Sen2Cor corresponding performance of 0.59. Focusing on the problem of optical satellite image scene classification, the main research contributions of this paper are: (a) an extended manually labeled Sentinel-2 database adding surface reflectance values to an existing dataset; (b) an ensemble-based and a Neural-Network-based ML models; (c) an evaluation of model sensitivity, biasness, and diverse ability in classifying multiple classes over different geographic Sentinel-2 imagery, and finally, (d) the benchmarking of the ML approach against the Sen2Cor package.