Search

In remote sensing, multiple input bands are derived from various sensors covering different regions of the electromagnetic spectrum. Each spectral band plays a unique role in land use/land cover characterization. For example, while integrating multiple sensors for predicting aboveground biomass (AGB) is important for achieving high accuracy, reducing the dataset size by eliminating redundant and irrelevant spectral features is essential for enhancing the performance of machine learning algorithms. This accelerates the learning process, thereby developing simpler and more efficient models. Our results indicate that compared individual sensor datasets, the random forest (RF) classification approach using recursive feature elimination (RFE) increased the accuracy based on F score by 82.86 % and 26.19 respectively. The mutual information regression (MIR) method shows a slight increase in accuracy when considering individual sensor datasets, but its accuracy decreases when all features are taken into account for all models. Overall, the combination of features from the Landsat 8, ALOS PALSAR backscatter, and elevation data selected based on RFE provided the best AGB estimation for the RF and XGBoost models. In contrast to the k-nearest neighbors (KNN) and support vector machines (SVM), no significant improvement in AGB estimation was detected even when RFE and MIR were used. The effect of parameter optimization was found to be more significant for RF than for all the other methods. The AGB maps show patterns of AGB estimates consistent with those of the reference dataset. This study shows how prediction errors can be minimized based on feature selection using different ML classifiers.

Accurate estimation of crop yield at the field scale plays a pivotal role in optimizing agricultural production and food security. Conventional studies have mainly focused on employing data-driven models for crop yield estimation at the regional scale, while large challenges may occur when attempting to apply these methods at the field scale. This is primarily due to the inherent complexity of obtaining reliable ground labels of yield for field validation, and the geographical independence and correlation that exists between fields. To effectively solve this problem, this study couples geographical, crop physiological knowledge and deep learning networks, and builds a graph-based deep learning framework by integrating high-medium spatial resolution active and passive remote sensing data (Sentinel-1, Sentinel-2 and Sentinel-3) and uses it to estimate field scale winter wheat yield. Firstly, a deep learning framework based on graph theory was constructed to achieve accurate estimation of field scale time series winter wheat growth parameter (Leaf Area Index, LAI), and then the growth mechanism of winter wheat and the specific factors affecting wheat yield formation were further considered, so as to improve the yield estimation accuracy of the traditional data-driven yield estimation model. Finally, the yield estimates of the proposed method were compared and analyzed for farmlands under different categories of agricultural disasters. The results showed that the graph-based two-branch network architecture (the Seq_Gra_Gd model) with the optimal meteorological data input strategy (meteorological data of the previous 15 d) had the optimal LAI estimation accuracy, and except for the jointing stage of winter wheat, the Seq_Gra_Gd model had a high and stable LAI estimation accuracy at the other main growth stages. The Seq_Gra_Gd model achieved good accuracy in estimating winter wheat yield (R2 = 0.73, RMSE = 590.43 kg·ha−1), and the introduction of the graph convolution module enabled the model to take into account the spatial distribution characteristics of stripe rust and lodging disasters well, which improved the yield estimation accuracy of affected winter wheat.

To further reveal the relationships between different variables and yield at each growth stage of winter wheat, an approach for estimating regional yields of winter wheat at multiple time scales was developed by assimilating the CERES-Wheat model simulations and remotely sensed observations. Specifically, the particle filter assimilation algorithm was chosen to assimilate the simulated soil moisture at the depth of 0–20 cm and leaf area index (LAI) and MODIS retrieved vegetation temperature condition index (VTCI) and LAI. The resonance periods of time series assimilated VTCIs and LAIs at different growth stages of winter wheat with crop yield were analyzed separately using the cross-wavelet transform to determine the variation relationships between the assimilated variables and yield at multiple time scales, and the calculated weights of assimilated VTCI and LAI at each growth stage of winter wheat were used to establish a yield estimation model. Both assimilated VTCI and LAI could comprehensively integrate the effects of the CERES-Wheat model simulations and remotely sensing observations, and cross-wavelet transformed time series VTCIs and LAIs at each growth stage had specific resonance periods with the time series yields, regardless of whether they were assimilated or not. Compared with the unassimilated variables, assimilated VTCI and LAI were given greater weights at the key growth stages, namely VTCI at the jointing and heading-filling stages and LAI at the heading-filling and milk maturity stages, enhancing feature extraction and the accuracy of yield estimation was improved.

In view of the ecological threat posed by cyanobacteria blooms in Taihu Lake (China), this paper presents a study on the area of cyanobacteria extent based on MODIS data using the quantum particle swarm optimization–random forest (QPSO-RF) machine learning algorithm. This paper selects multiple remote sensing input indices that can represent the characteristics of the primary underlying type in Taihu Lake. The proposed method performs best, with an F1 score of 0.91–0.98. Based on this method, the spatio-temporal variation of cyanobacteria blooms in the Taihu Lake complex was analyzed. During 2010–2022, the average area of cyanobacteria blooms in Taihu Lake increased slightly. Severe-scale cyanobacteria blooms occurred in 2015–2019. Cyanobacteria blooms were normally concentrated from May to November. However, the most prolonged extended duration occurred in 2017, lasting for eight months. Spatially, cyanobacteria blooms were mainly identified in the northwestern part of Taihu Lake, with an average occurrence frequency of about 10.0%. The cyanobacteria blooms often began to grow in the northwestern part of the lake and then spread to the Center of the Lake, and also dissipated earliest in the northwestern part of the lake. Our study is also beneficial for monitoring the growth of cyanobacteria blooms in other similar large lakes in long time series.

Study region: Poyang lake wetland Study focus: Evapotranspiration (ET) is a critical parameter in the hydrologic and energy budget, and it is important for understanding spatio-temporal variations in ET in large lake wetlands. Monthly ET in the Poyang Lake wetland during 2008–2017 was estimated by the remote sensing ET retrieval model based on a Nonparametric approach (RS-NP). And the spatio-temporal variation of ET and the influencing factors were analyzed. New hydrological insights for the region: The validation revealed the usefulness of the RS-NP model in a moist region and yielded a relative error of 6 % (12 %) in the wet (dry) season. Temporally, the average yearly ET was 884 mm with ET/P (precipitation) of 55 %− 65 %. The monthly ET/P ratio increased significantly from 40 %− 50 % from January to May to 90 %− 130 % from July to October. ET in June was lower by over 20 % than that in May and July with the lowest ET/P value (33 %). Spatially, relatively larger and lower ET occurred in the northern and southern parts, respectively. The main factors affecting ET variability were downwelling shortwave radiation (∼2/3 contribution) and water area (∼1/3 contribution) at the monthly scale. Precipitation and surface temperature dominantly controlled the hysteresis effects with a lag of one month and regulated LE monthly variations. This study improves our understanding of complicated water-atmosphere interactions and their linkages with environmental factors in large lake systems.

Regions with excessive cloud cover lead to limited feasibility of applying optical images to monitor crop growth. In this article, we built an upsampling moving window network for regional crop growth monitoring (UMRCGM) model to estimate the two key biophysical parameters (BPs), leaf area index (LAI), and canopy chlorophyll content (CCC) during the main growth period of winter wheat by using Sentinel-1 Synthetic Aperture Radar (SAR) and Sentinel-3 optical images. Sentinel-1 imagery is unaffected by cloudy weather and Sentinel-3 imagery has a wide width and short revisit period, the organic combination of the two will greatly improve the ability to monitor crop growth at a regional scale. The impact of two different types of SAR information (intensity and polarization) on the estimation of the two BPs was further analyzed. The UMRCGM model optimized the correspondence between inputs and outputs, it had more accurate LAI and CCC estimates compared with the three classical machine learning models, and had the highest accuracy at the green-up stage of winter wheat, followed by the jointing stage and the heading-filling stage, and the lowest accuracy was found at the milk maturity stage. The estimation accuracies of CCC were slightly higher than that of LAI for the first three growth stages of winter wheat, while lower than that of LAI for the milk maturity stage. This article proposes a new method for regional BPs (especially for CCC) estimation by combining SAR and optical imagery with large differences in spatial resolution under a deep learning framework.

Water stress is an important factor to be considered when using crop growth models for crop yields estimation. In this article, we propose a simple algorithm for yield estimate (SAFY) V model to estimate yields of winter wheat by integrating the time-series remotely sensed drought monitoring index, vegetation temperature condition index (VTCI), into SAFY model, and the fixed effective light-use efficiency parameter (elue0) in the SAFY model is modified to a new varied parameter (E) in the SAFY-V model. The parameter E can accurately describe the changes of water stress at the four growth stages of winter wheat and has a high correlation with the field measured yields (R2 values are 0.28, 0.34, 0.31 and 0.31, respectively). The SAFY-V model integrating the time series leaf area index (LAIs) and VTCIs which has the highest accuracy on dry aerial mass estimation compared with the SAFY model and SAFY-WB model (a combination of the SAFY model and water balance model), can better alleviate the yield underestimation and overestimation and greatly improve the estimation accuracy of winter wheat yield especially in rain-fed farmlands (R2 = 0.48, MAE = 1.05 t/ha and NMAE = 15.6%), and the accuracy of winter wheat yields estimates at the county scale was also satisfactory (R2 = 0.49, MAE = 0.73 t/ha and NMAE = 16.1% for five years). The proposed SAFY-V model of this article has few model parameters and low computational cost, which provides a significant reference for crop yield estimation at a regional scale.

The rapid and effective acquisition of crop yield information is critical to the stability of food markets and development and implementation of related policies. It is an important baseline observation that is used for ensuring regional and global food security. In this study, a novel deep learning framework was developed for winter wheat yield estimation using meteorological data and two remotely sensed indices, Vegetation Temperature Condition Index (VTCI) and Leaf Area Index (LAI) at the main growth stages of winter wheat in the Guanzhong Plain. The proposed deep learning model was based on Long Short-Term Memory (LSTM) neural network with an attention mechanism (ALSTM), which the main idea is to assign attention to the key parts of the input sequence that affect the target vectors so that the specific features can be accurately extracted. The ALSTM model provided an improved estimation accuracy (R2 = 0.63, MAPE = 8.20%, RMSE = 502.71 kg/ha, NRMSE = 11.15%) as compared with the LSTM (R2 = 0.55, MAPE = 13.46%, RMSE = 699.92 kg/ha, NRMSE = 15.52%). A validation based on leave-one-year-out-validation further substantiated the robustness of ALSTM with smaller values of NRMSE and MAPE (13.63% and 11.54%). We demonstrated that the ALSTM model provided good generalization ability for sampling sites under different farming systems, including irrigation and rain-fed sampling sites. In addition, we evaluated the relative importance of each input variable in determining yields based on stepwise sensitivity analysis. It was found that LAI at the heading-filling stage and the milk stage as well as VTCI at the jointing stage contributed more than other input feature variables towards the corresponding yield. In conclusion, our findings highlighted that the attention mechanism helped to improve the interpretability of neural networks and the ALSTM model along with remotely sensed biophysical indices can provide a reliable and robust estimation of crop yield. An accurate estimation of wheat yield is not only helping towards informed crop management decisions but it will improve efficiency and sustainability of farming operations.

The vegetation temperature condition index (VTCI) has been shown to perform well for drought monitoring using multiyear advanced very high-resolution radiometer (AVHRR) and moderate resolution imaging spectroradiometer (MODIS) data. Compared to single-year VTCI, the VTCI calculated by multiyear data can quantitatively reflect the degree of drought and precipitation in a region. Sentinel-3 is a recently launched remote sensing satellite with high temporal resolution and is similar to the satellites carrying AVHRR and MODIS sensors. One year of Sentinel-3 data is available for calculating the VTCI, and given the need for developing quantitatively drought monitoring capabilities, the aim of this study is to investigate the methods used to calculate the potential multiyear Sentinel-3 VTCI for quantitative drought monitoring. This is based on a comparison with multiyear Terra MODIS VTCI and a correlation analysis with cumulative precipitation data. The analysis results indicate that the potential multiyear Sentinel-3 VTCI can be accurately calculated from the single-year Sentinel-3 VTCI based on the linear correlation between the single-year VTCI and multiyear VTCI derived from Terra MODIS, which do not exhibit obvious systematic deviations from the multiyear Terra MODIS VTCI (the absolute value of the average bias 2 = 0.731, P

The tropical savanna in Brazil known as the Cerrado covers circa 23% of the Brazilian territory, but only 3% of this area is protected. High rates of deforestation and degradation in the woodland and forest areas have made the Cerrado the second-largest source of carbon emissions in Brazil. However, data on these emissions are highly uncertain because of the spatial and temporal variability of the aboveground biomass (AGB) in this biome. Remote-sensing data combined with local vegetation inventories provide the means to quantify the AGB at large scales. Here, we quantify the spatial distribution of woody AGB in the Rio Vermelho watershed, located in the centre of the Cerrado, at a high spatial resolution of 30 metres, with a random forest (RF) machine-learning approach. We produced the first high-resolution map of the AGB for a region in the Brazilian Cerrado using a combination of vegetation inventory plots, airborne light detection and ranging (LiDAR) data, and multispectral and radar satellite images (Landsat 8 and ALOS-2/PALSAR-2). A combination of random forest (RF) models and jackknife analyses enabled us to select the best remote-sensing variables to quantify the AGB on a large scale. Overall, the relationship between the ground data from vegetation inventories and remote-sensing variables was strong (R2 = 0.89), with a root-mean-square error (RMSE) of 7.58 Mg ha−1 and a bias of 0.43 Mg ha−1.

In Southeast Asia land use change associated with forest loss and degradation is a major source of greenhouse gas (GHG) emissions. This is of particular concern where deforestation occurs on peat soils. A business-as-usual (BAU) land change model was developed using Dinamica EGO© for a REDD+ Demonstration Activity area in south-east Jambi Province, Sumatra, Indonesia containing Berbak National Park (NP). The model output will be used as baseline land change predictions for comparison with alternative land cover management scenarios as part of a REDD+ feasibility study. The study area is approximately 376,000 ha with approximately 50% on peat soils. The model uses published 2000 and 2010 land cover maps as input and projects land cover change for thirty years until 2040. The model predicted that under a BAU scenario the forest area, 185,000 ha in 2010, will decline by 37% by 2040. In protected forest areas, approximately 50% of the study area, forest cover will reduce by 25%. Peat swamp forest will reduce by almost 37%. The greatest land cover category increases are plantation/regrowth areas (which includes oil palm) and open areas which each increase by 30,000 ha. These results indicate that the site has great potential as an Indonesian REDD+ Demonstration Activity.

This study provides a comparative analysis of two Sentinel-1 and one Sentinel-2 burned area (BA) detection and mapping algorithms over 10 test sites (100 × 100 km) in tropical and sub-tropical Africa. Depending on the site, the burned area was mapped at different time points during the 2015−2016 fire seasons. The algorithms relied on diverse burned area (BA) mapping strategies regarding the data used (i.e., surface reflectance, backscatter coefficient, interferometric coherence) and the detection method. Algorithm performance was compared by evaluating the detected BA agreement with reference fire perimeters independently derived from medium resolution optical imagery (i.e., Landsat 8, Sentinel-2). The commission (CE) and omission errors (OE), as well as the Dice coefficient (DC) for burned pixels, were compared. The mean OE and CE were 33% and 31% for the optical-based Sentinel-2 time-series algorithm and increased to 66% and 36%, respectively, for the radar backscatter coefficient-based algorithm. For the coherence based radar algorithm, OE and CE reached 72% and 57%, respectively. When considering all tiles, the optical-based algorithm provided a significant increase in agreement over the Synthetic Aperture Radar (SAR) based algorithms that might have been boosted by the use of optical datasets when generating the reference fire perimeters. The analysis suggested that optical-based algorithms provide for a significant increase in accuracy over the radar-based algorithms. However, in regions with persistent cloud cover, the radar sensors may provide a complementary data source for wall to wall BA detection.

Frequent cloud cover in the tropics significantly affects the observation of the surface by satellites. This has enormous implications for current approaches that estimate greenhouse gas (GHG) emissions from fires or map fire scars. These mainly employ data acquired in the visible to middle infrared bands to map fire scars or thermal data to estimate fire radiative power and consequently derive emissions. The analysis here instead explores the use of microwave data from the operational Sentinel-1A (S-1A) in dual-polarisation mode (VV and VH) acquired over Central Kalimantan during the 2015 fire season. Burnt areas were mapped in three consecutive periods between August and October 2015 using the random forests machine learning algorithm. In each mapping period, the omission and commission errors of the unburnt class were always below 3%, while the omission and commission errors of the burnt class were below 20% and 5% respectively. Summing the detections from the three periods gave a total burnt area of ∼1.6 million ha, but this dropped to ∼1.2 million ha if using only a pair of pre- and post-fire season S-1A images. Hence the ability of Sentinel-1 to make frequent observations significantly increases fire scar detection. Comparison with burnt area estimates from the Moderate Resolution Imaging Spectroradiometer (MODIS) burnt area product at 5 km scale showed poor agreement, with consistently much lower estimates produced by the MODIS data-on average 14%–51% of those obtained in this study. The method presented in this study offers a way to reduce the substantial errors likely to occur in optical-based estimates of GHG emissions from fires in tropical areas affected by substantial cloud cover.

Vegetation cover can affect the lithological mapping capability of space- and airborne instruments because it obscures the spectral signatures of the underlying geological substrate. Despite being widely accepted as a hindrance, few studies have explicitly demonstrated the impact vegetation can have on remote lithological mapping. Accordingly, this study comprehensively elucidates the impact of vegetation on the lithological mapping capability of airborne multispectral data in the Troodos region, Cyprus. Synthetic spectral mixtures were first used to quantify the potential impact vegetation cover might have on spectral recognition and remote mapping of different rock types. The modeled effects of green grass were apparent in the spectra of low albedo lithologies for 30%–40% fractional cover, compared to just 20% for dry grass cover. Lichen was found to obscure the spectra for 30%–50% cover, depending on the spectral contrast between bare rock and lichen cover. The subsequent impact of vegetation on the remote mapping capability is elucidated by considering the outcomes of three airborne multispectral lithological classifications alongside the spectral mixing analysis and field observations. Vegetation abundance was found to be the primary control on the inability to classify large proportions of pixels in the imagery. Matched Filtering outperformed direct spectral matching algorithms owing to its ability to partially unmix pixel spectra with vegetation abundance above the modeled limits. This study highlights that despite the limited spectral sampling and resolution of the sensor and dense, ubiquitous vegetation cover, useful lithological information can be extracted using an appropriate algorithm. Furthermore, the findings of this case study provide a useful insight to the potential capabilities and challenges faced when utilizing comparable sensors (e.g., Landsat 8, Sentinel-2, WorldView-3) to map similar types of terrain.

Image-based line segment extraction plays an important role in a wide range of applications. Traditional line segment extraction algorithms focus on the accuracy and efficiency, without considering the integrity. Serious line segmentation fracture problems caused by image quality will result in poor subsequent applications. To solve this problem, a multi-constrained line segment extraction method, based on multi-scale image space, is presented. Firstly, using Gaussian down-sampling with a classical line segment detection method, a multi-scale image space is constructed to extract line segments in each image scale and all line segments are projected onto the original image. Then, a new line segment optimization and purification strategy is proposed with the horizontal and vertical distances and angle geometric constraint relationships between line segments to merge fracture line segments and delete redundant line segments. Finally, line segments with adjacent positions are optimized using the grayscale constraint relationship, based on normalized cross-correlation similarity criterion for realizing the second optimization of fracture line segments. Compared with mainstream line segment detector and edge drawing lines methods, experimental results (i.e., indoor, outdoor, and aerial images) indicate the validity and superiority of our proposed methods which can extract longer and more complete line segments.

Time-series of imagery acquired by the Moderate Resolution Imaging Spectroradiometer (MODIS) has previously been used to estimate woody and herbaceous vegetation cover in savannas. However, this is challenging due to the mixture of woody and herbaceous plant functional types with specific contributions to the phenological signal and variations in soil background reflectance signatures observed from satellite. These factors cause variations in the accuracy and precision of woody cover estimates from different modelling approaches and datasets. Here, woody cover is estimated over Kruger National Park (KNP) from the MODIS 16-day composite time-series data using dry season NDVI/SAVI images and applying NDVIsoil determination methods. The woody cover estimates when NDVIsoil was ignored had R2 = 0.40, p < 0.01, slope = 1.01, RMSE (root mean square error) = 15.26% and R2 = 0.32, p < 0.03, slope = 0.79, RMSE = 16.39% for NDVIpixel and SAVIpixel, respectively, when compared to field plot data of plant functional type fractional cover. The woody cover estimated from the soil determination methods had a slope closer to 1 for both NDVI and SAVI but also a slightly higher RMSE. For a soil-invariant method, RMSE = 19.04% and RMSE = 17.34% were observed for NDVI and SAVI respectively, while for a soil-variant method, RMSE = 18.28% and RMSE = 19.17% were found for NDVI and SAVI. The woody cover estimated from all models had a high correlation and significant relationship with LiDAR/SAR based estimates and a woody cover map produced by Bucini. Woody cover maps are required for vegetation succession monitoring, grazing impact assessment, climate change mitigation and adaptation research and dynamic vegetation model validation.

Land surface temperature (LST) is a crucial parameter in the interaction between the ground and the atmosphere. The Sentinel-3A Sea and Land Surface Temperature Radiometer (SLSTR) provides global daily coverage of day and night observation in the wavelength range of 0.55 to 12.0 μm. LST retrieved from SLSTR is expected to be widely used in different fields of earth surface monitoring. This study aimed to develop a split-window (SW) algorithm to estimate LST from two-channel thermal infrared (TIR) and one-channel middle infrared (MIR) images of SLSTR observation. On the basis of the conventional SW algorithm, using two TIR channels for the daytime observation, the MIR data, with a higher atmospheric transmittance and a lower sensitivity to land surface emissivity, were further used to develop a modified SW algorithm for the nighttime observation. To improve the retrieval accuracy, the algorithm coefficients were obtained in different subranges, according to the view zenith angle, column water vapor, and brightness temperature. The proposed algorithm can theoretically estimate LST with an error lower than 1 K on average. The algorithm was applied to northern China and southern UK, and the retrieved LST captured the surface features for both daytime and nighttime. Finally, ground validation was conducted over seven sites (four in the USA and three in China). Results showed that LST could be estimated with an error mostly within 1.5 to 2.5 K from the algorithm, and the error of the nighttime algorithm involved with MIR data was about 0.5 K lower than the daytime algorithm.

In recent decades petroleum pollution in the tropical rainforest has caused significant environmental damage in vast areas of the Amazon region. At present the extent of this damage is not entirely clear. Little is known about the specific impacts of petroleum pollution on tropical vegetation. In a field expedition to the Ecuadorian Amazon over 1100 leaf samples were collected from tropical trees in polluted and unpolluted sites. Plant families were identified for 739 of the leaf samples and compared between sites. Plant biodiversity indices show a reduction of the plant biodiversity when the site was affected by petroleum pollution. In addition, reflectance and transmittance were measured with a field spectroradiometer for every leaf sample and leaf chlorophyll content was estimated using reflectance model inversion with the radiative tranfer model PROSPECT. Four of the 15 plant families that are most representative of the ecoregion (Melastomataceae, Fabaceae, Rubiaceae and Euphorbiaceae) had significantly lower leaf chlorophyll content in the polluted areas compared to the unpolluted areas. This suggests that these families are more sensitive to petroleum pollution. The polluted site is dominated by Melastomataceae and Rubiaceae, suggesting that these plant families are particularly competitive in the presence of pollution. This study provides evidence of a decrease of plant diversity and richness caused by petroleum pollution and of a plant family-specific response of leaf chlorophyll content to petroleum pollution in the Ecuadorian Amazon using information from field spectroscopy and radiative transfer modelling.

Synthetic Aperture Radar (SAR) signals respond to the interactions of microwaves with vegetation canopy scatterers that collectively characterise forest structure. The sensitivity of S-band (7.5–15 cm) backscatter to the different forest types (broadleaved, needleleaved) with varying aboveground biomass (AGB) across temperate (mixed, needleleaved) and tropical (broadleaved, woody savanna, secondary) forests is less well understood. In this study, Michigan Microwave Canopy Scattering (MIMICS-I) radiative transfer model simulations showed strong volume scattering returns from S-band SAR for broadleaved canopies caused by ground/trunk interactions. A general relationship between AirSAR S-band measurements and MIMICS-I simulated radar backscatter with forest AGB up to nearly 100 t/ha in broadleaved forest in the UK was found. Simulated S-band backscatter-biomass relationships suggest increasing backscatter sensitivity to forest biomass with a saturation level close to 100 t/ha and errors between 37 t/ha and 44 t/ha for HV and VV polarisations for tropical ecosystems. In the near future, satellite SAR-derived forest biomass from P-band BIOMASS mission and L-band ALOS-2 PALSAR-2 in combination with S-band UK NovaSAR-S and the joint NASA-ISRO NISAR sensors will provide better quantification of large-scale forest AGB at varying sensitivity levels across primary and secondary forests and woody savannas.

Remote sensing-based approaches to lithological mapping are traditionally pixel-oriented, with classification performed on either a per-pixel or sub-pixel basis with complete disregard for contextual information about neighbouring pixels. However, intra-class variability due to heterogeneous surface cover (i.e., vegetation and soil) or regional variations in mineralogy and chemical composition can result in the generation of unrealistic, generalised lithological maps that exhibit the “salt-and-pepper” artefact of spurious pixel classifications, as well as poorly defined contacts. In this study, an object-based image analysis (OBIA) approach to lithological mapping is evaluated with respect to its ability to overcome these issues by instead classifying groups of contiguous pixels (i.e., objects). Due to significant vegetation cover in the study area, the OBIA approach incorporates airborne multispectral and LiDAR data to indirectly map lithologies by exploiting associations with both topography and vegetation type. The resulting lithological maps were assessed both in terms of their thematic accuracy and ability to accurately delineate lithological contacts. The OBIA approach is found to be capable of generating maps with an overall accuracy of 73.5% through integrating spectral and topographic input variables. When compared to equivalent per-pixel classifications, the OBIA approach achieved thematic accuracy increases of up to 13.1%, whilst also reducing the “salt-and-pepper” artefact to produce more realistic maps. Furthermore, the OBIA approach was also generally capable of mapping lithological contacts more accurately. The importance of optimising the segmentation stage of the OBIA approach is also highlighted. Overall, this study clearly demonstrates the potential of OBIA for lithological mapping applications, particularly in significantly vegetated and heterogeneous terrain.

Assessments of forest cover, forest carbon stocks and carbon emissions from deforestation and degradation are increasingly important components of sustainable resource management, for combating biodiversity loss and in climate mitigation policies. Satellite remote sensing provides the only means for mapping global forest cover regularly. However, forest classification with optical data is limited by its insensitivity to three-dimensional canopy structure and cloud cover obscuring many forest regions. Synthetic Aperture Radar (SAR) sensors are increasingly being used to mitigate these problems, mainly in the L-, C- and X-band domains of the electromagnetic spectrum. S-band has not been systematically studied for this purpose. In anticipation of the British built NovaSAR-S satellite mission, this study evaluates the benefits of polarimetric S-band SAR for forest characterisation. The Michigan Microwave Canopy Scattering (MIMICS-I) radiative transfer model is utilised to understand the scattering mechanisms in forest canopies at S-band. The MIMICS-I model reveals strong S-band backscatter sensitivity to the forest canopy in comparison to soil characteristics across all polarisations and incidence angles. Airborne S-band SAR imagery over the temperate mixed forest of Savernake Forest in southern England is analysed for its information content. Based on the modelling results, S-band HH- and VV-polarisation radar backscatter and the Radar Forest Degradation Index (RFDI) are used in a forest/non-forest Maximum Likelihood classification at a spatial resolution of 6 m (70% overall accuracy, κ = 0.41) and 20 m (63% overall accuracy, κ = 0.27). The conclusion is that S-band SAR such as from NovaSAR-S is likely to be suitable for monitoring forest cover and its changes.

Radar backscatter from forest canopies is related to forest cover, canopy structure and aboveground biomass (AGB). The S-band frequency (3.1–3.3 GHz) lies between the longer L-band (1–2 GHz) and the shorter C-band (5–6 GHz) and has been insufficiently studied for forest applications due to limited data availability. In anticipation of the British built NovaSAR-S satellite mission, this study evaluates the benefits of polarimetric S-band SAR for forest biophysical properties. To understand the scattering mechanisms in forest canopies at S-band the Michigan Microwave Canopy Scattering (MIMICS-I) radiative transfer model was used. S-band backscatter was found to have high sensitivity to the forest canopy characteristics across all polarisations and incidence angles. This sensitivity originates from ground/trunk interaction as the dominant scattering mechanism related to broadleaved species for co-polarised mode and specific incidence angles. The study was carried out in the temperate mixed forest at Savernake Forest and Wytham Woods in southern England, where airborne S-band SAR imagery and field data are available from the recent AirSAR campaign. Field data from the test sites revealed wide ranges of forest parameters, including average canopy height (6–23 m), diameter at breast-height (7–42 cm), basal area (0.2–56 m2/ha), stem density (20–350 trees/ha) and woody biomass density (31–520 t/ha). S-band backscatter-biomass relationships suggest increasing backscatter sensitivity to forest AGB with least error between 90.63 and 99.39 t/ha and coefficient of determination (r2) between 0.42 and 0.47 for the co-polarised channel at 0.25 ha resolution. The conclusion is that S-band SAR data such as from NovaSAR-S is suitable for monitoring forest aboveground biomass less than 100 t/ha at 25 m resolution in low to medium incidence angle range.