Ground Samples Research Articles

Crop Residue Cover Percentage Estimation from RGB Images Using Transfer Learning and Ensemble Ordinal Regression

Highlights A transfer learning strategy improved residue estimates from high-resolution RGB imagery. The best method used probabilistic estimates of expert classifiers to estimate residue cover. This research confirms the utility of RGB imagery to quantify residue cover in agricultural fields. Abstract. Plant residue on the soil surface increases the sustainability of food and fiber production in agricultural systems. Automated assessments of residue cover based on imagery have the potential to reduce labor and human bias associated with in-field measurements. We evaluate the capacity of a transfer learning strategy to improve the determination of residue level from high-resolution RGB images. The imagery for the study was collected from 88 field locations in 40 row crop fields in five Missouri counties between mid-April and early July in 2018 and 2019. At each field location, 50 contiguous 0.3 m × 0.2 m region of interest (ROI) images (ground sampling distance of 0.014 cm pixel--1) were extracted from imagery, resulting in a dataset of 4,400 ROI images; 3,000 were used for cross-validation and training (data collected in 2018) and 1,400 were used for testing (data collected in 2019). The percentage residue for each ROI image (ground truth) was determined by a bullseye grid method (n = 100). Features were extracted from ROI images using the VGGNet-16 CNN model, a pre-trained convolutional neural network model. We extracted 1,472 features per ROI using a global averaging and pooling strategy. The optimum feature set was identified using recursive feature elimination using a support vector machine (RFE-SVM). To estimate crop residue percentage using selected features, expert two-class SVMs were trained to separate adjacent levels of residue cover, where the rationale of the ensemble was to allow each of the two-class SVMs to find the hyperplanes that maximize the margin between the corresponding two consecutive classes. Based on the distance of the samples to these hyperplanes, probabilistic estimates of the data-point belonging to the class were computed. With the combined knowledge of probabilistic estimates from each expert classifier, the percentage crop residue cover of each ROI image was calculated. We tested our approach with 3-, 4-, 5-, and 8-class problems, achieving the best results with the 8-class problem with r2 = 0.93 at the ROI level, r2 = 0.97 at the field-location level, and minimal bias in residue estimates in low residue conditions. These results are superior to other reported estimates of percent residue derived from imagery. This research confirms the utility of high-resolution RGB imagery to quantify residue cover in agricultural systems. Keywords: Convolutional neural network, Soil erosion, Support vector machine, Transfer learning.

Pressure-induced white-light emission of a Cu(i) cluster with effective blue phosphorescence

Blue phosphorescence with the PLQY of 68.5% is achieved in a Cu(i) cluster by dispersing the phosphor centres. And the ground sample show white-light emission and the CIE is (0.34, 0.36), which is further fabricated as the WLED with the Ra of 81.5.

Limitations and future perspectives for satellite-based soil carbon monitoring

Soil organic carbon (SOC) plays a crucial role in terrestrial C storage and ecosystem services. Agricultural management practices have the potential to increase C inputs and reduce its losses. However, uniform standard protocols for measuring, monitoring, and assessing changes using remote sensing is lacking for SOC in the scientific literature. In this discussion paper, we present techniques for collecting and analyzing ground samples and employing remote sensing to quantify SOC, along with its limitations and future perspectives. Our analysis identified a number of key limitations to advancing the science for remotely sensed terrestrial C in croplands including i) lack of consensus in sampling depth and density, ii) the absence of a standard (or universally accepted) laboratory procedure and statistical methodology, and iii) lack of details on imagery pre-processing or information on the spectral properties of the targeted soils. Establishing standard protocols for ground-truth data collection and remote sensing approaches, as well as a knowledge of the impacts of diverse soil types, land uses, and landscapes on C assessment, are all required to enhance the accuracy and reliability of future SOC assessments.

Employment of RPA (Remotely Piloted Aircraft) in Identifying Imperfections in the Asphalt Pavement of the Transoeste Bus Rapid Transit (BRT) System

The Brazilian Road network is constantly used by the population and its maintenance is an essential charge to ensure the performance of the pavement. Therefore, the verification of asphalt conditions is extremely relevant to analyze their structural and functional qualities. It is noteworthy that this verification is usually performed by technicians manually being the results slow and often with impalpable and partial analysis. Given this, we see the need to study the benefits of obtaining this information indirectly without the need for inloco man. In this work we will focus on the use of Remotely Piloted Aircraft System to obtain the images that will enable the analysis of asphalt conditions. The place where the study will take place is the corridor of BRT Transoeste in Rio de Janeiro, which connects the region of Barra da Tijuca to Santa Cruz and Campo Grande, passing through Guaratiba and Recreio dos Bandeirantes. The following procedures will be adopted: Flight Planning with RPAS observing GSD Ground Sample Distance advocated by the scientific literature, field support using GNSS NTRIP technology, processing of images obtained with the use of Agisoft software and generating Orthorectified Orthophoto. Finally, the main objective of the work is to verify the potential of the metric quality of the images from RPAS in identifying damage to the asphalt surface

Analysis of Physico-Chemical Characteristics and Heavy Metals Content in Water Samples Utilizing Atomic Absorption Spectrophotometry in Kafin Hausa, Jigawa State, Nigeria.

Heavy metals contamination of ground, stream, and river water ecosystems is a worldwide environmental problem (Sekabira et al., 2010), and between the wide diversity of contaminants affecting water resources, heavy metals receive particular concern considering their strong toxicity even at low concentrations (Marcovecchio et al., 2007). The levels of heavy metals in surface, ground, and tap water samples collected in Kafin Hausa were determined using Atomic Absorption spectrophotometry, whereas nitrate, ammonium, and phosphate ions were analysed using spectrophotometric methods. The following ranges were obtained for Mn (0.032±0.001 to 0.092±0.001 mg/l), Fe (0.029±0.001 to 0.051±0.001 mg/l), Pb (0.017±0.001 to 0.0095±0.001 mg/l), Cu (0.016±0.001 to 0.032±0.001 mg/l), Cd (0 to 0.0055±0.001 mg/l), Zn (0.016±0.001 to 0.032±0.001 mg/l), NO3-N (1.16±0.261 to 1.55±0.232 mg/l), PO4-P (0.596±0.267 to 1.26±0.502 mg/l) and NH4-N (0.700±0.041 to 0.771±0.094 mg/l), pH (8.2±0.1 to 8.4±0.1), EC (134.7±0.5 to 161.6±0.4 µS/cm), Temperature (26.5±0.00 to 27.8±0.00 oC) and TDS (64.9±0.3 to 80.8±0.2 mg/l). The results of this study showed that out of those metals and anions analyzed only that of PO43 exceeded the permissible limit of good portable water set by NAFDAC/WHO (2011) in all the samples analyzed.

Hy-TeC: a hybrid vision transformer model for high-resolution and large-scale mapping of canopy height

Accurate and timely monitoring of forest canopy height is critical for assessing forest dynamics, biodiversity, carbon sequestration as well as forest degradation and deforestation. Recent advances in deep learning techniques, coupled with the vast amount of spaceborne remote sensing data offer an unprecedented opportunity to map canopy height at high spatial and temporal resolutions. Current techniques for wall-to-wall canopy height mapping correlate remotely sensed information from optical and radar sensors in the 2D space to the vertical structure of trees using lidar's 3D measurement abilities serving as height proxies. While studies making use of deep learning algorithms have shown promising performances for the accurate mapping of canopy height, they have limitations due to the type of architectures and loss functions employed. Moreover, mapping canopy height over tropical forests remains poorly studied, and the accurate height estimation of tall canopies is a challenge due to signal saturation from optical and radar sensors, persistent cloud cover, and sometimes limited penetration capabilities of lidar instruments. In this study, we map heights at 10 m resolution across the diverse landscape of Ghana with a new vision transformer (ViT) model, dubbed Hy-TeC, optimized concurrently with a classification (discrete) and a regression (continuous) loss function. This model achieves significantly higher accuracy than previously employed convolutional-based approaches (ConvNets) optimized with only a continuous loss function. Hy-TeC results show that our proposed discrete/continuous loss formulation significantly increases the sensitivity for very tall trees (i.e., > 35 m). Overall, Hy-TeC has significantly reduced bias (0.8 m) and higher accuracy (RMSE = 6.6 m) over tropical forests for which other approaches show poorer performance and oftentimes a saturation effect. The height maps generated by Hy-TeC also have better ground sampling distance and better sensitivity to sparse vegetation. Over these areas, Hy-TeC showed an RMSE of 3.1 m in comparison to a reference dataset while the baseline ConvNet model had an RMSE of 4.3 m. Hy-TeC, which was used to generate a height map of Ghana using free and open access remotely sensed data with Sentinel-2 and Sentinel-1 images as predictors and GEDI height measurements as calibration data, has the potential to be used globally.

Mechanism of Potassium Release from Feldspar by Mechanical Activation in Presence of Additives at Ordinary Temperatures.

To improve the potassium availability of feldspar at ordinary temperatures, the mechanical grinding and addition of sodium hydroxide/salts were employed to study the effects of mechanical activation and strong alkali addition on particle characteristics, water-soluble potassium, and the available potassium of feldspar. A laser particle size analyzer was utilized for the direct determination of particle size distribution (PSD) using ground samples. The Brunauer-Emmett-Teller (BET) method was employed for specific surface areas. X-ray diffraction (XRD) was employed for structural characterization, scanning electron microscopy (SEM) for morphology exploration, and energy dispersive spectroscopy (EDS) to determine the chemical composition of potassium feldspar powder. The results revealed that the mechanical activation of potassium feldspar could reduce the particle size and produce agglomerated nanoparticles in the later period. The addition of NaOH and sodium salt did not cause agglomeration, and NaOH dissolved the nanoparticles. The water-soluble potassium content of feldspar in each treatment increased during mechanical grinding, from 21.64 mg kg-1 to 1495.81 mg·kg-1, by adding NaOH 5% weight of potassium feldspar powder and to 3044.08 mg·kg-1 by adding NaOH 10% weight with effects different from those of mechanical shaking. By comparison, only 162.93 mg·kg-1 water-soluble potassium was obtained by adding NaOH 5% weight. The dissolved potassium in the former case was significantly higher than in the latter, and the addition of NaOH and sodium salts significantly enhanced the water-soluble potassium contents due to ion exchange. Furthermore, the addition of sodium hydroxide improved the water-soluble potassium due to its mechanochemical action on potassium feldspar. The mechanical energy changed the crystal structure of potassium feldspar, explaining the increase in available potassium. The addition of sodium salts did not promote change in the feldspar's structure, thereby did not raise the available potassium content. The reason for this was related to the mechanochemical action on sodium hydroxide and feldspar, which could promote the dissolution of fine particles, thereby incrementing the available potassium.

Genetic Programming Guided Mapping of Forest Canopy Height by Combining LiDAR Satellites with Sentinel-1/2, Terrain, and Climate Data

Accurately mapping the forest canopy height is vital for conserving forest ecosystems. Employing the forest height measured by satellite light detection and ranging (LiDAR) systems as ground samples to establish forest canopy height extrapolation (FCHE) models presents promising opportunities for mapping large-scale wall-to-wall forest canopy height. However, despite the potential to provide more samples and alleviate the stripe effect by synergistically using the data from two existing LiDAR datasets, Global Ecosystem Dynamics Investigation (GEDI) and Ice, Cloud, and land Elevation Satellite-2 (ICESat-2), the fundamental differences in their operating principles create measurement biases, and thus, there are few studies combining them for research. Furthermore, previous studies have typically employed existing regression algorithms as FCHE models to predict forest canopy height, without customizing a model that achieves optimal performance based on the current samples. These shortcomings constrain the accuracy of predicting forest canopy height using satellite LiDAR data. To surmount these difficulties, we proposed a genetic programming (GP) guided method for mapping forest canopy height by combining the GEDI and ICESat-2 LiDAR data with Sentinel-1/2, terrain, and climate data. In this method, GP autonomously constructs the fusion model of the GEDI and ICESat-2 datasets (hereafter GIF model) and the optimal FCHE model based on the explanatory variables for the specific study area. The outcomes demonstrate that the fusion of the GEDI and ICESat-2 data shows high consistency (R2 = 0.85, RMSE = 2.2m, pRMSE = 11.24%). The synergistic use of the GEDI and ICESat-2 data, coupled with the optimization of the FCHE model, substantially improves the precision of forest canopy height predictions, and finally achieves R2, RMSE, and pRMSE of 0.64, 3.38m, and 16.08%, respectively. In summary, our research presents a reliable approach to accurately estimate forest canopy height using remote sensing data by addressing measurement biases between the GEDI and ICESat-2 data and overcoming the limitations of traditional FCHE models.

Source apportionment of methane emissions from the Upper Silesian Coal Basin using isotopic signatures

Abstract. Anthropogenic emissions are the primary source of the increase in atmospheric methane (CH4) levels. However, estimates of anthropogenic CH4 emissions still show large uncertainties at global and regional scales. Differences in CH4 isotopic source signatures δ13C and δ2H can help to constrain different source contributions (e.g., fossil, waste, agriculture). The Upper Silesian Coal Basin (USCB) represents one of the largest European CH4 emission regions, with more than 500 Gg CH4 yr−1 released from more than 50 coal mine ventilation shafts, landfills, and wastewater treatment plants. During the CoMet (Carbon Dioxide and Methane Mission) campaign in June 2018 methane observations were conducted from a variety of platforms including aircraft and cars to quantify these emissions. Besides the continuous sampling of atmospheric methane concentration, numerous air samples were taken from inside and around the ventilation shafts (1–2 km distance) and aboard the High Altitude and Long Range Research Aircraft (HALO) and DLR Cessna Caravan aircraft, and they were analyzed in the laboratory for the isotopic composition of CH4. The airborne samples downwind of the USCB contained methane from the entire region and thus enabled determining the mean signature of the USCB accurately. This mean isotopic signature of methane emissions was -50.9±0.7 ‰ for δ13C and -226±9 ‰ for δ2H. This is in the range of previous USCB studies based on samples taken within the mines for δ13C but more depleted in δ2H than reported before. Signatures of methane enhancements sampled upwind of the mines and in the free troposphere clearly showed the influence of biogenic sources. We determined the source signatures of individual coal mine ventilation shafts using ground-based samples. These signatures displayed a considerable range between different mines and also varied for individual shafts from day to day. Different layers of the USCB coal contain thermogenic methane, isotopically similar to natural gas, and methane formed through biogenic carbonate reduction. The signatures vary depending on what layer of coal is mined at the time of sampling. Mean shaft signatures range from −60 ‰ to −42 ‰ for δ13C and from −200 ‰ to −160 ‰ for δ2H. A gradient in the signatures of subregions of the USCB is reflected both in the aircraft data and in the ground samples, with emissions from the southwest being most depleted in δ2H and emissions from the south being most depleted in δ13C, which is probably associated with the structural and lithostratigraphic history of the USCB and generation and migration processes of methane in the coal. The average signature of -49.8±5.7 ‰ in δ13C and -184±32 ‰ in δ2H from the ventilation shafts clearly differs from the USCB regional signature in δ2H. This makes a source attribution using δ2H signatures possible, which would not be possible with only the δ13C isotopic signatures. We assume that the USCB plume mainly contains fossil coal mine methane and biogenic methane from waste treatment, because the USCB is a highly industrialized region with few other possible methane sources. Assuming a biogenic methane signature between and −320 ‰ and −280 ‰ for δ2H, the biogenic methane emissions from the USCB account for 15 %–50 % of total emissions. The uncertainty range shows the need of comprehensive and extensive sampling from all possible source sectors for source apportionment. The share of anthropogenic–biogenic emissions of 0.4 %–14 % from this densely populated industrial region is underestimated in commonly used emission inventories. Generally, this study demonstrates the importance of δ2H-CH4 observations for methane source apportionment in regions with a mix of thermogenic and biogenic sources and, especially in our case, where the δ13C signature of the coal mine gas has a large variability.

An LC-MS/MS Method for the Simultaneous Analysis of 380 Pesticides in Soybeans, Kidney Beans, Black Soybeans, and Mung Beans: The Effect of Bean Grinding on Incurred Residues and Partitioning.

The significance of sample grinding is frequently disregarded during the development of analytical methods, which are often validated with spiked samples that may not accurately reflect incurred residues. This study investigated the particle size of ground beans as a key factor in optimizing extraction efficiency in order to develop a simple quick, easy, cheap, effective, rugged, and safe (QuEChERS)-based modified method for identifying 380 pesticides in beans using liquid chromatography-tandem mass spectrometry. The efficacy of pesticide extraction was found to be significantly affected by particle size. With small particle sizes (>40 mesh), no supernatant was recovered after QuEChERS partitioning. Therefore, a simple modification was performed before partitioning. The modified method was validated for selective extraction of pesticides, limits of quantification, linearity, accuracy, and precision. This method is simple to implement and, therefore, useful for the analysis of pesticide residues in beans.

ON THE ACCURACY OF YOLOV8-CNN REGARDING DETECTION OF HUMANS IN NADIR AERIAL IMAGES FOR SEARCH AND RESCUE APPLICATIONS

Abstract. The use of deep learning techniques especially in conjunction with convolutional neural networks (CNN) has attracted major attention of the remote sensing community. Main use cases are object detection, image classification and image segmentation. The paper will focus on object detection, specifically on detection of humans. In search and rescue applications it is common to map larger areas with downward facing cameras. However, there are many training data sets for CNNs showing oblique images which strongly differ from nadir aerial images used for real-time maps.To circumnavigate this issue, an unique data set was created. It solely contains nadir images at different ground sample distances (GSD) varying from one to five centimetres. Diversity of the training data is ensured through various flights using an unmanned aerial vehicle (UAV) at different locations. GSD dependency is valuable prior knowledge as it enhances the difficulty associated with human detection in aerial images. An image, depicting a human at one centimetre GSD contains much more information than the same human depicted in an image of three centimetres. That is one reason why networks trained on a variety of ground sample distances possibly struggle to detect humans reliably on a certain GSD. The unique data set consists of four subsets (divided by GSD). Each subset contains 1000 manually annotated humans, augmented by rotation and colour shift resulting in 12000 training samples used to train the new released YoloV8 CNN. The entire training and test process is unified to ensure comparable input conditions.

A new approach to surveying cliff-dwelling endangered plants using drone-based nap-of-the-object photography: A case study of Clematis acerifolia

Cliffs are essential refuges for endangered plants, but their inaccessibility and inadequate reliable survey methods make the assessment of the survival status of these plants difficult. We employed the drone-based Nap-of-the-Object Photography (NoOP) technique, which involves flying close to an object's surface to automatically capture high-precision images of the object surface, to investigate an endangered cliff plant, Clematis acerifolia, in North China. A drone (DJI M300 RTK) equipped with a visible-light mapping camera (DJI Zenmuse P1) was used for automated NoOP flights, enabling the collection and reconstruction of millimeter-level, high-precision three-dimensional (3D) models of the cliff plots. Across 12 cliffs, we identified and measured the precise coordinates and values of eight microtopographical and five morphological characteristic factors for 1444 C. acerifolia plants. The Ground sampling distance (GSD) of the high-precision 3D models reached 0.10 cm/pixel, and the mean value of the root mean squared error of the sampled positioning coordinates was 4.65 cm/pixel. Correlation and correspondence analyses were used for data analyses. The results indicated that the growth status of C. acerifolia was significantly influenced by a complex factor interaction, particularly plant altitude, crevice width, and crevice length. This method for cliff plant surveys using drone-based NoOP determines the parameters for such surveys and demonstrates the effectiveness and potential of this method for cliff plant investigations. The primary microtopographical factors affecting the growth and distribution status of C. acerifolia were identified, providing pointers for translocation conservation and propagation reintroduction of this endangered plant.

State-wide forest canopy height and aboveground biomass map for New York with 10 m resolution, integrating GEDI, Sentinel-1, and Sentinel-2 data

Investigating the quantity of forest aboveground biomass (AGB) is crucial for understanding the role forests play in the global carbon cycle. The canopy height model (CHM) is a critical component in estimating AGB, as it provides a three-dimensional representation of the tree canopy. Traditional CHM estimation methods are time-consuming, labor-intensive, and expensive, particularly at large-scales. Remote sensing is a cost-effective and efficient alternative approach, providing valuable information over large areas in a timely manner. The Global Ecosystem Dynamics Investigation (GEDI) onboard the International Space Station is a space-based light detection and ranging (LiDAR) system designed to collect information on vertical structures of vegetation. One major problem with the collection of GEDI data is that it provides limited information over discrete ground samples, also known as footprints, and thus do not provide wall-to-wall gridded height products. The objective of this study was twofold: a) to integrate the GEDI LiDAR footprint heights with Sentinel-2 multispectral imagery to generate a 10 m wall-to-wall CHM map of New York State (NYS), USA for the year 2019 and b) to improve our previously generated AGB map (both accuracy and resolution) of NYS for the year 2019 by fusing Sentinel-2 multispectral, Sentinel-1 synthetic aperture radar (SAR), and the produced CHM. To generate the 10 m CHM map, the GEDI footprints height measurements were extrapolated using Sentinel-2 imagery and a random forest model. The CHM that was produced was assessed by using GEDI footprints that were not part of the training phase and were therefore independent (extrapolated). Comparing our 10 m CHM with the available global 30 m CHM map provided by Potapov et al. (2021) over NYS shows significant improvement not only in terms of spatial resolution, but also in terms of accuracy. The root mean square error (RMSE) of our 10 m CHM is 4.4 m while this value is 7.49 m for the 30 m CHM over NYS. Similarly, the R2 value for the 10 m CHM map is 0.74, while that of the 30 m CHM is 0.46. Finally, the integration of produced 10 m CHM, Sentinel-1, and Sentinel-2 datasets were utilized to create a 10 m AGB map of NYS with the RMSE of 39.49 Mg/ha, and R2 of 0.65. The results demonstrate the potential of integrating GEDI, Sentinel-1, and Sentinel-2 data for providing a valuable tool for large-scale mapping of forest canopy structure and biomass, which can help to inform forest management and carbon accounting efforts.

Obtaining cumin (Cuminum Cymminum) essential oils by steam distillation. Part I: investigation of operation parameters on essential oil yield and distribution of oil composition

AbstractBACKGROUNDThe effect of pretreatment (grinding), packing material (raschig rings) and vacuum on the yield and the component distribution of the essential oil of cumin during steam distillation was investigated. Concurrently, kinetic studies were conducted to examine the essential oil yield.RESULTSAt the end of 120 min, the yield under normal conditions was 0.79, 1.26, 1.90 and 2.36 for nonground and 20‐, 40‐ and 80‐s ground cumin seeds, respectively. Equivalent figures in a vacuum were 0.64, 1.56 and 1.95 for nonground, 40‐ and 80‐s ground cumin seeds, respectively. When raschig rings were used, the yield value was 1.99 and 2.06 under normal and vacuum conditions, respectively, for cumin seeds ground for 80 s. Under normal conditions, the cumin aldehyde percentages of the nonground sample, and the 40‐ and 80‐s ground samples was 51.31, 33.59 and 32.54, respectively, and under vacuum, 70.00, 49.02 and 40.59, respectively. Six kinetic models were applied with sufficient R2 values. Drug release models were first applied to kinetic data successfully.ConclusionGrinding cumin seeds in a blender increased the essential oil yield up to three‐fold and Raschig rings as a packing material also contributed to this increase. The effect of complex mechanisms such as sample positioning and flow regime on the essential oil yield during the washing stage was observed. Although there was minimal difference (3.8%) in the yield values for the experiments with nonground cumin samples, the slight variation in essential oil yield observed in the experiments with each ground cumin sample may be attributed to the particle distribution, positioning and flow regime. So, five cases were determined depending on positioning. Cumin aldehyde has the highest composition in all time intervals after 10 min at the fractional distillation under vacuum. However sabinene was relatively higher only in the early time interval at 0‐10 min. © 2023 The Authors. Journal of Chemical Technology and Biotechnology published by John Wiley & Sons Ltd on behalf of Society of Chemical Industry (SCI).

Efficacy evaluation of apple sunburn mitigation techniques in WA 38 cultivar using crop physiology sensing system

This study evaluated the efficacy of two heat stress mitigation techniques, i.e., modified evaporative cooling (EC), and drape/shade net (SN) against untreated control (UC) in terms of fruit surface temperature (FST) suppression and internal fruit quality for cv. WA 38. The EC treatment used foggers (flow rate: 4.1 L s−1 ha−1) spaced at 3.3 m along the row to create mist within the canopy. The experiment was conducted in a 0.2 acre block planted on two randomly distributed rootstocks [Geneva 41 (G41) and Malling 9-Nic 29 (M9)], trained in a bi-axis vertical wall architecture. Treatments were divided into three sub-blocks. Alternate wet and dry zones were observed in the sub-block treated with fogger-based EC and was therefore subdivided into EC-Wet and EC-Dry treatments. The proximal thermal-RGB imaging and in-field weather sensing integrated crop physiology sensing system (CPS2.0) was upgraded with edge computing capability to monitor real-time apple FST, an indicator of heat stress. Deployed nodes monitored imagery and weather data based FST in each treatment at an interval of 5-min throughout the growing season (i.e., during fruit maturation). Aerial imagery-based canopy temperature mapping with ground sampling distance of 4 cm/pixel was also utilized to monitor the canopy heat stress at key growth stages. CPS2.0 imagery derived apple FST was effectively reduced by EC-Wet (G41: 27.1 ± 0.4 and M9: 28.8 ± 0.4 °C) (mean ± std. error) and SN (G41: 29.8 ± 0.2 and M9: 30.5 ± 0.1 °C) treatments compared to the UC (G41: 33.1 ± 0.2 and M9: 31.4 ± 0.4 °C) throughout the production season. However, convective cooling from EC-Dry (G41: 33.1 ± 0.2 and M9: 32.5 ± 0.2 °C) treatment failed to mitigate heat stress. Results suggest that fogger spacing, and orientation need adjustments along tree-rows for effective sunburn management. Overall, canopy temperature quantified by aerial imagery was significantly lower under modified EC (30.3 ± 1.3 °C) than UC (34.2 ± 1.3 °C) treatment. Compared to UC, EC-Wet and SN protection enhanced fruit quality with higher or similar crop load, fruit firmness, and quality.

A cross-resolution transfer learning approach for soil moisture retrieval from Sentinel-1 using limited training samples

Synthetic Aperture Radar (SAR) data is increasingly popular as a data source for global near-surface soil moisture mapping, but large-scale applications are still challenging due to the complex scattering process and the cumbersome data preprocessing. The emergence of deep learning methods has allowed advances in the remote sensing of large-scale surface parameters, but its application in SAR soil moisture retrieval has suffered from the availability of ground soil moisture measurements. Accordingly, this study proposed a cross-resolution transfer learning framework, with the assumption that sophisticated models for different spatial resolutions share a similar model architecture and trainable parameters. A robust high-resolution model can thus be trained with fewer samples by using coarse models. Accordingly, 25 deep learning models were pre-trained taking ∼387,000 Soil Moisture Active Passive (SMAP) Level-3 9 km enhanced passive soil moisture measurements as the truth, with an average validation RMSE of 0.03 m3/m3. They were then transferred to finer grids of 0.1–1 km using a small number of in-situ samples. A total of ∼190,000 daily soil moisture measurements from the international soil moisture network (ISMN) were used to evaluate the proposed framework in three scenarios. The results show that 1) 5000–6000 random samples are sufficient to achieve a target RMSE of 0.06 m3/m3; 2) training samples from a short period (2 or 4 months for Sentinel-1) of 2021 resulted in an overall RMSE of ∼0.068 m3/m3 in an independent period of 2016–2020; 3) the transfer learning also improved the retrieval accuracy (10–30% in relative) over areas without ground samples used for training but failed to yield an acceptable accuracy over mountainous areas. The promising results from this study confirmed the effectiveness of using “pre-trained models + scenario specific models” for regional to global soil moisture retrieval from Sentinel-1.

FUSION OF DIRECT GEOREFERENCED AERIAL DRONE WITH TERRESTRIAL LASER SCANNER DATA THE CASE OF THE ROMAN BATHS OF AMATHUS, CYPRUS

Abstract. The fusion of geomatic techniques with different accuracy and resolution has been used in recent years to applications of geometrical documentation for several archaeological sites. In this research article, the case of the Roman Baths of Amathus in Limassol, Cyprus was investigated.Aerial photogrammetry and Terrestrial Laser Scanning (TLS) techniques were used to produce the final product. The photogrammetric data were based on Post-Processed Kinematic (PPK) imagery while the TLS georeference was based on targets.The main research objective of this paper is to examine the reliability of PPK photogrammetric data without the use of Ground Control Points (GCPs) and how well they can be integrated with TLS data. Also, the two acquisition techniques were compared and indicated that 0.005 m Ground Sampling Distance (GSD) resolution aerial images and 0.0061/10 m resolution scans can be qualitatively fused.The research paper will present the entire methodology up to the generation of the final 3D photorealistic product.

DRONE-BASED CROP TYPE IDENTIFICATION WITH CONVOLUTIONAL NEURAL NETWORKS: AN EVALUATION OF THE PERFORMANCE OF RESNET ARCHITECTURES

Abstract. This study investigates the application of deep learning techniques, specifically ResNet architectures, to automate crop type identification using remotely sensed data collected by a DJI Mavic Air drone. The imagery was captured at an altitude of 30 meters, maintaining an average airspeed of 5 m/s, and ensuring a front and side overlap of 75% and 65%, respectively. The pre-flight planning and image acquisition was facilitated through the Drone Deploy platform, yielding a dataset consisting of 1488 aerial photographs covering the study area. These images possess an average ground sampling distance (GSD) of 22.2 millimetres. The dataset was meticulously labelled with "maize" and employed to train three distinct ResNet architectures, namely ResNet-50, ResNet-101, and ResNet-152. The evaluation of these models was based on accuracy and processing time. Notably, ResNet-50 emerged as the most proficient, achieving an accuracy rate of 82% with a precision score of 0.5 after just two hours of initial training, while ResNet-101 and ResNet-152 architectures achieved 27% and 24% accuracy, respectively. These outcomes underscore the potential of ResNet-50 architecture, even with a limited dataset, as a valuable tool for precise crop-type classification within the precision agriculture domain.

The Classification Characteristics and Dynamic Changes of Desert Vegetation Based on Unmanned Aerial Vehicle Remote Sensing

The common ground sample method of vegetation survey requires a lot of manpower and time. Meanwhile, many vegetation information may not be revealed because the scale of standard sample is usually only 100 m2. For this reason, this study developed a set of desert vegetation sample survey method based on UAV technology, and further developed a set of vegetation spatial distribution pattern analysis method based on it, and carried out an empirical study with typical shrub communities in Inner Mongolia as an example. It is expected to provide scientific basis for wind and sand control, vegetation restoration and ecological construction in the ecologically fragile area at the edge of Inner Mongolia.

Assessment of Selected Ground Samples Around a Sugar Factory, Shirol, Maharashtra, India

Ground water quality surrounding a sugar factory was studied for the impact of effluent discharged by the industry by applying multivariate statistical techniques. Although the samples recorded the EC values well within the permissible range and a close assessment of the analysis and the evaluated statistical approach disclosed that the overall quality of ground water was suitable for domestic purpose. Concentrations of cations Na+, K+, Ca+, and Mg+ varied from 67 to 43, 2.2 to 1.1, 70 to 106, and 25 to 37 ppm, respectively. Na+, Ca2+, Mg2+, and K+ were major cations in chronological order accompanied by HCO3-, Cl-, and SO42- for the major anions. Na+ was the most widespread cation and K+ was the lowest abundant, whereas bicarbonates were the most common and SO4 was the least significant cation under anions. More than 95% of the samples fall under Doneen's Class I category, and every sample demonstrated SAR values below 10, which indicated high suitability for applying water domestic as well as for crops.