This PDF file contains the front matter associated with SPIE Proceedings Volume 12262, including the Title Page, Copyright information, Table of Contents, and Conference Committee listings.

The EU H2020 I-Seed project aims for sustainable environmental monitoring of topsoil and air above soil environments by employing Unmanned Aerial Vehicles (UAV) to distribute, localize and read-out of the fluorescence signal of the artificial I-seeds. Reaction with relevant environmental parameters and process of bio-degradation will induce a change of fluorescence in the artificial seeds, which will be recorded from an airborne platform with sufficient signal-to-noise ratio to identify the concentration of targeted soil parameters, such as mercury, carbon dioxide, humidity and temperature. Remote sensing based laser-induced fluorescence systems are used in atmospheric and environmental monitoring, where the emitted fluorescence is collected at a working distance of couple of meters to hundreds of meters from the zone of interest. However, technology maturation, miniaturization and cost has always been a major bottleneck for developing mini-UAV based active spectroscopic systems. Here we present the design ideas and results of first lab-scale experiments to realize an active laser-induced fluorescence system on UAV platform. Such a system has potential to address not only the sustainable environment monitoring and agricultural production, but also the threats in food security, climate change and sustainable resource management.

Maize (Zea mays L., corn) crops are extensively used in food and biofuel production worldwide. A number of protocols have been proposed to use leaf color as an indicator of the health status of maize plants. Color perception is a complex process, however. The correct interpretation of its outcomes depends on several aspects. Accordingly, a variety of spectral vegetation indices have also been proposed to monitor the development of these plants. These indices usually require a number of spectral reflectance and transmittance samples taken from selected specimens using specialized sensors. Since these radiometric quantities do not depend neither on the spectra of the light sources nor on the physiological characteristics of the human visual system, these indices are not subject to color perception issues. The visual feedback provided by the chromatic attributes of plant leaves, on the other hand, can enable a broader assessment of the net effect of several environmental factors affecting an entire maize crop. Also, these attributes can be obtained using spectral reflectance and transmittance samples already employed in the computation of the aforementioned indices. These aspects indicate the potential benefits of the combined use of vegetation indices and leaf chromatic attributes in the monitoring of maize crops. Ideally, one would like to employ a number of spectral samples that would maximize the color fidelity to sensor costs ratio. In this paper, we address this practical trade-off. More specifically, using hyperspectral reflectance and transmittance data for maize specimens, we performed colorimetric experiments to obtain a lower bound for the number of spectral reflectance and transmittance samples sufficient to achieve a high degree of fidelity in the reproduction of maize leaves’ colors under distinct illumination conditions.

The increasing scarcity of freshwater supplies, notably due to drought conditions elicited by climate changes, has a detrimental impact in the yield of crops worldwide. Not surprisingly, this type of abiotic stress factor has been the object of extensive studies in plant physiology, precision agriculture and remote sensing fields, just to name a few. Its effects on C₄ species like corn (Zea mays L., maize) tend to follow a cumulative pattern with distinct stages of severity. In its early, moderate, stage, it has a minor impact on the plants’ chlorophyll contents. Nonetheless, it is essential to detect and manage it before irreversible damage to the plants’ photosynthetic apparatus can occur. The unifacial corn leaves are equipped with stress adaptation mechanisms such as the rearrangement of their chloroplasts. Accordingly, their chromatic attributes can change under moderate water deficit conditions. These changes, however, are often subtle and their assessment can be impaired by varying illumination conditions. Alternatively, multispectral vegetation indices can be employed to assist the monitoring of these plants’ water status. In this paper, we address these aspects. More specifically, we investigate the sensitivity of corn leaves’ chromatic attributes to changes in their optical properties in response to moderate water stress. Furthermore, we propose an alternative index to assist the monitoring of these changes in the visible (photosynthetic) spectral domain. We carried out our investigation using a first-principles in silico experimental approach supported by measured data. Our findings are expected to contribute to the advance of the current understanding about moderate water stress related effects on C₄ plants. Such an understanding, in turn, is instrumental for the development of new water-stress monitoring technologies with a higher reliability to cost ratio.

Phenocams that capture images of a given area in the RGB or near-infrared (NIR) spectrum have been used for more than a decade to estimate phenology in natural landscapes and crop fields. The aim of our study is to estimate phenological parameters, start (SOS) and end (EOS) of season, for barley, from RGB and NIR Phenocam and compare them with in-situ observations from two sites, one with growing season 2014/2015 and the other with growing season 2021/2022. Time series of Phenocam Green Chromatic Coordinate (GCC) and Normalized Difference Vegetation Index (NDVI) were computed then scaled to Harmonized Landsat-8 and Sentinel-2 surface (HLS), available for both sites, and Sentinel-2 (S2), available for only one site, datasets. The HLS and S2 datasets were gap filled with classical and machine learning methods before the scaling. Phenological parameters were extracted from the scaled GCC and NDVI Phenocam data and from the gap filled HLS and S2 datasets. Our preliminary results show that the SOS can be modelled with one day difference compared with the in-situ observed with the scaled Phenocam NDVI and a week difference compared with the in-situ observed with gap filled HLS and S2 datasets with both vegetation indices.

Contemporary climate change has induced an urge to monitor and manage the forest ecosystems to maintain the carbon and energy balance of the system. For understanding the behavior of a forest, its productivity is the key. Monitoring a forest in terms of its productivity provides a holistic understanding of the response of the forest to external stimuli in form of climate change as well as change in its physiological processes. Productivity depends on the effectiveness of the various interlinked physiological processes indicating the climatic, anthropogenic, and geochemical influence on the biosphere. Himalayan ecosystem, one the most fragile ecosystems with home to large biodiversity, requires a clear understating of forest and its response to climatic variability. For assessing these responses, we conducted a study to examine the trend of net primary productivity (NPP) from 2002-2021 across the Western Himalayan region. Remotely sensed (NPP) data for the last 20 years at 500m spatial resolution have been acquired from Moderate Resolution Imaging Spectroradiometer (MODIS) using Google Earth Engine platform. All the analysis was carried out using R software and maps have been generated using ArcGIS software. The study area witnessed a positive trend throughout the study period. The annual average NPP increased from 3.18 Mg C ha^-1yr^-1 to 4.98 Mg C ha^-1yr^-1 from 2002-2021. Examining the spatial variations, we found NPP has increased in 88% of the total study area, whereas the rest of the area has experienced a decrease. Disintegrating the trend into different slope categories indicated that 88.22% of the area under 15°-30° slope show positive trend in NPP followed by 0-15 and 30-60 both having 86% area with a positive trend. Analysis revealed that a 15°-30° slope is most favorable for forest growth in terms of NPP.

Assessment of crop development provides important information for agricultural management and yield prediction. Accurate estimates of crop phenology require information about actual planting dates and hydrometeorological variables such as soil moisture and temperature. In 2019, an unseasonably wet spring across Illinois due to persistent flood events delayed or prevented normal planting, ultimately reducing crop production. A study was conducted to develop a novel approach to determine day of planting for corn using remotely sensed soil moisture and temperature. Planting dates were determined based on soil moisture conditions and soil temperature suitable for corn germination. Estimated planting dates were then used to initiate accumulating growth degree days derived from air temperature to estimate the crop progress stages. Results were evaluated by comparing the estimated planting schedule and timelines of crop phenology to those obtained from USDA’s Crop Progress Report. The estimated planting schedule shows that more than 90 % of the fields were planted by June 2 in normal seasons, while only about 45 % of fields were planted by that date in 2019, similar to that observed from the USDA’s report. Overall, estimates of accumulative planted areas over the planting season for both normal seasons and for 2019 match up well to USDA’s report. Nearly two weeks of average planting delay for 2019, as compared to normal seasons, resulted in 20-30 days of delays for all growth stages through the season, which can also be observed from both estimated and reported timelines of crop phenology. The RMSDs between estimated and reported crop progress timelines over the growing seasons for 2015-2020 and 2019 are 6.0 and 4.8 days, respectively. These results demonstrate the feasibility of utilizing soil moisture and temperature to estimate the planting schedule and crop phenology for agricultural management. This proposed approach is particularly applicable for assessment of planting schedule during extremely wet soil conditions. In addition, the study provides an opportunity to estimate planting dates along with the sowing season that can be further applied with weather forecasting data for crop yield prediction and assessment in the early stage of the growing season.

The Portuguese agri-forestry system Montado occupies about 730,000 hectares, which is about 8% of total area of Portugal. The maintenance of this biodiverse and complex land cover system is threatened, among other causes, due to frequent tillage to manage shrubs encroachment. In order to characterize Montado areas, we develop a neural network algorithm for identifying regions with trees, shrubs, covered and/or bare soil in grasslands. For this purpose, we used high-resolution RGB orthophotos (spatial resolution of 25 cm) that cover mainland Portugal. They were collected during the summer and autumn of 2018. The labelling of the used images was performed through an unsupervised method (Gaussian mixtures), which was validated through visual interpretation. The deep convolutional neural networks architecture used was U-net, which has been used in the literature to segment remote sensing images with a high performance. To train models, 800 orthophotos with 10,000 m2 each were used. They were divided between training and test set. A hyperparameter tuning was performed, namely the number of filters, dropout rate, batch size and the training/test partition percentage. In the best model, the overall classification performance (measured on the test set) was 89%, the recall 90% and the mean intersection of the union of 79%. Nevertheless, identification of shrubs had the lowest performance (accuracy of 85%), which are mainly confused with trees that have similar spectral signature. This model enables the identification of the status of Montado ecosystem regarding shrub encroachment for better future management.

Dams play a significant role in the storage, supply, and capitalization of water resources. This study analyzes the influences of land-use and climate factors on two interconnected dams, Gaborone and Bokaa dams, in the semi-arid Botswana from 2001 to 2019. Using Random Forest regression (RFR) and Vector AutoRegression (VAR) models, the monthly dam water levels were predicted based on the variabilities of rainfall and temperature, climate indices (DSLP, Aridity Index (AI), SOI and Niño 3.4) and land-use land-cover (LULC) information comprising of built-up, cropland, water, forest, shrubland, grassland and bare-land. The prediction results using the climate factors and climate indices show that for both dams, RFR was able to detect the correlations between the dam water levels with R2 of between 0.805 and 0.845 with min, average and max temperatures as the best combined predictors. Using differenced stationary datasets, VAR identified the climate indices as the suitable predictors for water levels in Gaborone and Bokaa dams with R2 of 0.929 and 0.916 respectively. VAR also detected LULC to be strongly correlated to the dam water levels. Nevertheless, LULC was considered as more significant when combined with the climate-based predictor variables. Comparatively, VAR was able to detect the interdependence between the two dams and with the other conjunctive water sources as the water levels in both dams were not significantly correlated with rainfall trends, while RFR relied on the seasonal temperature variabilities to accurately predict the fluctuations in the dam water levels.

The Energy Union Framework Strategy is pushing the entire world to move from fossil fuels to renewable energy to tackle climate changes and mitigate their effects. Among the clean energy alternatives, the sun is recognized as the most abundant and inexhaustible source and the energy production can be carried out through photovoltaic panels. Nevertheless, such a solar park requires the use of large land areas, stolen, in such a way, from food production, which demand has strongly increased in the last few years due to the growing world population. Thus, agrophotovoltaic systems, also known as agrivoltaic structures, are under way to meet the above-mentioned needs synergistically. This has led to the necessity of monitoring solar panels amount and allocation. Their detection is challenging since, albeit their spectral signature is totally different from that one emitted from other land covers, their occurrence received little attention in the field of remote sensing. Thus, in this study, a proper rule-based model for distinguishing photovoltaic panels developed on eCognition environment was proposed. Such a model is based on the combination of Object-Based Image Analysis and machine learning algorithm. Indeed, after optimizing segmentation parameters and analyzing morphological features of the panels, the Random Forest classification algorithm was implemented. Lastly, classification accuracy was evaluated. The experimentation was conducted on the study area of Viterbo (Lazio Region, Italy) by adopting open medium-resolution satellite data (Sentinel 2). This research showed promising results in classifying targets for almost all months of the time series, except for the months of October and November where there is a lowering of the accuracy value due to the variability of spectral signatures.

Portugal produced a land cover map for 2018 based on Sentinel-2 data and represents 13 classes, including agriculture, six tree forest species, and shrubland. The map was updated for 2020. The strategy focused on three strata where annual changes occur: S1 (agriculture) due to crop rotation, S2 (forest and shrubland) due to wildfires and clear-cuts, and S3 (fire scars and clear-cuts of previous years) where vegetation regeneration occurs. The methodology included i) change detection, ii) classification, and iii) knowledge-based rules. Stratum S1 was classified with images of the entire 2020 crop year and a training dataset extracted from the national Land Parcel Identification Systems (LPIS) of 2020. The land cover nomenclature was expanded and class agriculture was split in three distinct classes, hence resulting a map with 15 classes in total. Change detection, implemented in stratum S2, analyzed the profile of NDVI since 2018 to find potential loss of vegetation. S2 and S3 were classified through two stages. First, images of the entire 2020 crop year were used and then data of October 2020 (end of crop year) to capture late changes. The training points of the 2018 land cover map were used, but only if not associated with NDVI change. For all the three strata, knowledge-based rules corrected misclassifications and ensured consistency between the maps. A comparison between 2018 and 2020 reveal important land cover dynamics related to vegetation loss and regeneration on ~5% of the country.

This paper presents an annual crop classification exercise considering the entire area of continental Portugal for the 2020 agricultural year. The territory was divided into landscape units, i.e. areas of similar landscape characteristics for independent training and classification. Data from the Portuguese Land Parcel Identification System (LPIS) was used for training. Thirty-one annual crops were identified for classification. Supervised classification was undertaken using Random Forest. A time-series of Sentinel-2 images was gathered and prepared. Automatic processes were applied to auxiliary datasets to improve the training data quality and lower class mislabeling. Automatic random extraction was employed to derive a large amount of sampling units for each annual crop class in each landscape unit. An LPIS dataset of controlled parcels was used for results validation. An overall accuracy of 85% is obtained for the map at national level indicating that the methodology is useful to identify and characterize most of annual crop types in Portugal. Class aggregation of the annual crop types by two types of growing season, autumn/winter and spring/summer, resulted in large improvements in the accuracy of almost all annual crops, and an overall accuracy improvement of 2%. This experiment shows that LPIS dataset can be used for training a supervised classifier based on machine learning with high-resolution remote sensing optical data, to produce a reliable crop map at national level.

Detailed information on the extent and composition of tree species in a forest is crucial for scientific studies and forest management plans. In the recent years, remote sensing has emerged as a powerful tool in gathering different vegetation biophysical parameters. Further, the advancements in the state-of-art machine learning techniques have enhanced the process of combining multi-temporal, multi-sensor datasets, handling and analyzing large variable datasets to produce results with higher precision and accuracies. Ground-based census and mapping of tree species is a cumbersome, time taking and expensive exercise. Thus, Machine Learning Algorithm (MLA) based classification of remotely-sensed datasets has become a topical aspect of research interest. There are only a handful of studies mapping the distribution of forest tree species in the Western Himalayas using satellite imagery. The present study is amongst the pioneer studies undertaken in India to map one of the dominant tree species i.e. Pinus roxburghii, commonly known as Chir pine, by integrating machine learning techniques with the remote sensing technology. A supervised Random Forest MLA has been employed to discriminate between tree species based upon various spectral and topographical variables from Sentinel imagery. The methodology employed produced reliable maps of distribution of Chir pine forests in the study area, Uttarakhand, India with adequate accuracy. As the study has been undertaken using open and freely-available datasets such as Sentinel, Google Earth Engine (GEE) platform and ML libraries, it has the potential for adoption in countries like India for forestry research and inventory in a cost-effective manner. The methodology can be reproduced for delineation of different forest tree species to produce distribution maps with enhanced accuracies.

Quantum Machine Learning (QML) is a branch of quantum computing that combines classical machine learning with the principles of quantum mechanics. It is emerging as an alternative to classical machine learning which exploits the quantum mechanical properties of entanglement and superposition to express the hidden patterns in the data. This reduces computational resources also the time required for processing. This research work is a comparative study, which compares the overall performance of Classical multi-class Support Vector Classifiers (SVC) with Quantum multi-class Support Vector Classifiers (QSVC). In this work, we used benchmark Hyperspectral Remotely Sensed datasets namely, Pavia University and Salinas-A on IBM gate-based Quantum Computer(QC). Here, in QSVC, kernel is generated by QC, and Qiskit’s Support Vector Classifier is used for classification. Classification of the pixels into their respective classes was experimented using two techniques, One vs One (OVO) and One v/s Rest (OVR). Quantum kernels are very expressive when compared to their classical counterparts and can learn complex data more efficiently. The overall accuracy of classification by QSVC is comparable to that of the classical SVC. We summarize our research by saying that QSVC performs better than SVC.

In this study we map Demographic and Health Surveys (DHS) urban household water supply data from 30 African countries and 52 DHS-surveys to Sentinel 2 RGB data and show that modern convolutional neural networks can find a mapping function and predict abstract variables derived from DHS data, like household water supply. In addition, the purpose of this research is to show the ability of such networks to predict data for areas and countries where no survey data are available. Therefore, we use one-year medians of 2×2km cloud removed Sentinel 2 tiles at the surveyed locations in a VGG19 CNN and classify sources of water supply. In addition, we perform a regression analysis for the distance and the first principal component of a PCA. We achieve an F1-score of up to 0.76 for the classification and an r² of 0.76 for the prediction of the first principal component. The prediction of the distance to the water source is less precise with an r² of 0.57, which is potentially due to the extreme skewness of input data. In further studies, we want to prove that these results can also be achieved for rural and mixed models as well as for other food security indicators, such as asset wealth.

The Shaheen cyclone triggered coastal areas of Al-Batinah Governorate of the Sultanate of Oman and caused devastating impacts on vegetation areas, infrastructure and properties that resulted in severe damages and human casualties. A comprehensive evaluation of the cyclone is essential to identify the most impacted areas in the Governorate especially in its four regions Al-Musanaah, Al-Suwaiq, Al-Khaboura and Saham. An advanced techniques and very high resolution datasets have been used to study, analyze and mapping the effects caused by the shaheen Cyclone. The systematic approach included investigating changes before and after the cyclone of various parameters such as vegetation coverage, detection of buildup damages in agriculture lands, detailed study on coastline changes and inundations in agriculture areas & urban community. Both pre-classification and post classification change detection techniques were used to assess the impact of the cyclone. Using very high resolution datasets and application of latest techniques of Geographical information system and remote sensing like vegetation indices, deep learning models, spatial analysis and advanced object based detection methods were used to analyze the damages caused by the cyclone. Agricultural land change detection and its coverage calculation was studied and mapped. All individual vegetation parcels within the study area were analyzed and delineated. Date palm trees classification and counting was conducted and mapped. Inundations in agriculture lands and urban buildings in the agriculture areas were identified and mapped. The changes in the coastline and marine features were studied and mapped using latest object based classification. The outcome of this study was helpful in identifying the most affected areas and providing tempo-geospatially damage assessment that assist the humanitarian aid as well as paving the road for future hazard mitigation and new protection strategies.

Potato cultivation is regularly affected by Alternaria solani, a destructive foliar pathogen causing early blight, a premature defoliation of potato plants resulting in yield losses. Currently, Alternaria is treated through preventive application of chemical crop protection productions, following warnings based on weather predictions and visual observations. Automatic detection could make the mapping of early blight more accurate, reducing production losses and application of crop protection products. Current research explores the potential of deep learning of high resolution imagery within precision agriculture, mainly using supervised learning. However, available datasets are often limited in size and variation, which reduces the robustness of the developed models. Here, we present a convolutional network to detect Alternaria and evaluate the influence of sampling size, sampling balance and sampling accuracy on the model performance. These analyses are based on ultra-high-resolution datasets of modified RGB cameras obtained with unmanned aerial vehicles (UAV) and collected over experimental in-field Alternaria trials. By using this varied dataset instead of a single-time dataset, higher accuracies are achieved. The method is relatively robust for imbalances of the training dataset. Further, we show that labeling quality plays a role, but that an error of up of to 20% of labeling is acceptable for good results. In conclusion, extra variability leads to more robust disease detection, desirable for in-field application.

Forests are an integral part of the natural environment providing social, economic, and environmental benefits. Though storms are an important part of natural forest dynamics, large magnitude storms can lead to uprooting of trees, known as windthrows. Post-storm management relies on proper and fast detection of windthrows. In this work, we study the detection of windthrows due to storm David in the coniferous forests of southern Lower-Saxony, Germany as an image segmentation problem. Two deep learning methods, previously researched U-Nets and current state-of-the-art DeepLabv3+ are compared. Often storm damaged forests are surveyed many months later under good weather conditions, however, we study a winter storm surveyed in winter conditions 19 days after the storm. Moreover, we generate a detailed prediction map by segmenting the input scenery into four classes, namely, no forest, forest with no windthrows, forests with windthrows, and cleared areas. The data consists of four spectral channels and we study different 3-channel combinations and input image tile sizes to obtain the best configuration for windthrow detection. DeepLabv3+ is found to outperform U-Net with a prediction accuracy of 86.27% for windthrows, with best accuracy of 95.03% across all classes, and a class IoU of 0.7440 compared to a prediction accuracy of 78.66% and class IoU of 0.6892 for U-Nets. Deeplabv3+ was able to process 2048 × 2048 mosaics with input image tile size of 512 × 512 in nearly 889ms. Thus, a fast and well performing windthrow detection model based on DeepLabv3+ is developed.

This study analyses the influence of climate change on the melting of snow on the Chimborazo Mountain peak using remote sensing and a simpler mathematical model. Climate change is undoubtedly real, and one of its leading causes is increased carbon dioxide emissions caused by industrial activities. This climate change phenomenon manifests itself in several ways, such as temperature increase and precipitation variation. We studied the influence of these variables on a mountain peak in Ecuador called Chimborazo. Since Chimborazo Mountain has a considerable size, the area of its ice glacier is sufficiently large enough to be measured and studied. Estimating the glacier area on Chimborazo’s peak was carried out using photointerpretation over Satellite Remote Sensing in a GIS software, applying the best images without cloud per year of Landsat images from 1979 to 2020 because Ecuador has a high cloud density all year. The climate change data are collected from the Intergovernmental Panel on Climate Change (IPCC), matching the years of Remote Sensing data to construct an input dataset for the mathematical model. Then, in the RStudio, the Partial Least Square (PLS) model was executed, where it was determined how many of the combined variables (Climate change data) in the independent vector components could be used in the modelling. Thus, concluding that using the seven components explains 93% of the variation of the results of the area (Remote Sensing extracted area data). The results indicate that the maximum provincial temperature and CO2 country emissions are the variables with the most significant influence on the melting of snow on Chimborazo. Therefore, this melting is influenced by climate change. Additionally, a simulation based on the PLS model is used to compute the Chimborazo snow area until 2050. Thus, a PLS model and Remote Sensing variables can explain the climate change in the snow-capped Ecuadorian mountains in a first approach.

This paper presents a multitemporal analysis aimed at investigating the spatial and temporal consistency of Surface Soil Moisture (SSM), retrieved from optical satellite data by means of multispectral indexes, in a test area in central Italy. In particular, multitemporal series of images acquired by Landsat satellites were processed to obtain SSM estimates by using multispectral indices proposed in the literature. Retrieved SSM values and their temporal trends were compared in order to verify their consistency with the pluviometric data of the same period and the correlation between them. The results have highlighted the usefulness of multispectral indices for monitoring wide areas over long periods of time, yet with some factors, such as geology, pedology and phenology, that can affect the quality of the results and reduce the correlation with pluviometric trend.

The present study emphasize on the adequacy and advantage of remote sensing and GIS-based techniques to identify and assess the resource potential zones in Takoli Gad Watershed, India. Different thematic maps like elevation map, drainage map, soil map, geology map, slope map, rainfall map and land use/land cover map have been prepared using ASTER DEM, LISS IV and SOI topographical maps and suitable secondary data. The present study aimed to investigate various morphological characteristics by measuring linear, areal and relief aspects. The morphometric characterization indicate dynamic role in distinguishing the topographical and hydrological behaviour of the watershed. A composite map of the Resource Potential Zones was developed by integrated different thematic maps and morphometric characterization data base. In this study, geographical information system (GIS), remote sensing, weighted overlay analysis and analytical hierarchy process (AHP) methods have been used for mapping of available natural resources in the Takoli Gad watershed. The overlaid based on analytical hierarchical process weightage prioritization at a constituency ratio of 0.095. The resource potential zones categorizes into the excellent (1.59 %), good (25.61%), moderate (45.21 %), poor (26.41 %), and very poor (1.18 %) zones. The present study helps in the sustainable management of water, agriculture, forest and other natural resources to meet the basic minimum needs of people, and improve their socio-economic conditions. Decision and policy makers can use the information gathered from such case study for long term development plan of watershed.

Agricultural runoff and municipal sewage generate excessive nutrient input in estuaries, which disturbs the ecosystem's natural balance. Most monitoring programs require in situ measurements, which are expensive, time-consuming, and lack spatial and temporal resolution. Extensive research focuses on mitigating these costs by minimizing the indicators or using remote sensing tools. One of the currently investigated options is the application of unmanned aerial vehicles (UAVs) data since it can narrow the multi-resolution gap between the in situ and satellite data. As an initial step of such a multi-scaling approach, we focused on testing the applicability of existing algorithms developed for the Sentinel-2 multispectral data (MS) on our hyperspectral (HS) data obtained using UAV. We applied the available algorithms to estimate three water quality (WQ) parameters: Chlorophyll a (Chl a), Colored Dissolved Organic Matter (CDOM), and turbidity (TUR), for the in situ data acquired at the estuary of the River Jadro near the city of Split (Croatia). The higher spectral resolution obtained by HS imaging enabled us to use the specific wavelengths corresponding to the satellite bands for which the initial algorithms were developed. Moreover, we made one synthetic dataset of MS data, obtained by spectral resampling of HS data using spectral response functions for Sentinel 2 sensors given by ESA. By using these corresponding bandwidths, the initial study found medium and poor correlations with the WQ parameters: Chl a (R²=0.48), turbidity (R²=0.07), and CDOM (R²=0.22). Furthermore, all algorithms revealed higher correlations when using HS data compared to synthesized MS data. However, to fortify these results, we need to test more algorithms and compare the results with satellite reflectance data. Moreover, the future goals of this study are to develop new algorithms which could serve as surrogate data for satellite predictions.

Active Thermography (AT) is a well-known non-contact and non-invasive imaging technique that has gained great interests in agriculture in recent years. It has been used to evaluate physical and physiological characteristics of plants such as: transpiration rates, heat capacity of the leaves, local water content, response to UV interaction and it fits well with emerging demands of the precision agriculture management strategy. According to this technique, the surface of the sample under investigation is stimulated using an external heat source and its thermal response is detected and recorded using infrared camera. Different strategies can be used for both the measurement protocol and for data analysis. Copper has been widely used in agriculture as a fungicide and bactericide for many decades. Applied on leaf, copper based fungicide (CBF) remains deposited and it is not absorbed into plant tissues, causing accumulation problems that needs to be monitored and controlled, also by using modern technologies. In this work, we test and compare different AT methods to detect and to monitor the presence of CBF on leaves. Our experimental results demonstrate that methodological approaches based on AT can be used to engineer effective remote tools to evaluate in real-time the presence of copper on plants, allowing a tentative of quantification and, therefore, to optimize its use in the agricultural practices.

GIS models of the empirical methods Blaney-Criddle, Hammon, De Bruin and Makkink have been developed in ArcMap model-builder environment to estimate daily mean reference evapotranspiration (ETo). MODIS LST Day and Night products along with meteorological data were used as inputs. The application was on Peloponnese, Greece, for months December and August of years 2016-2019. For the validation of the models FAO PenmanMonteith (FAO-PM) was used as a reference. For the evaluation of the models Root Mean Square Error (RMSE), Mean Bias (MB) and Normalized mean Bias (NMB) were computed. Hargreaves-Samani’s and Hansen’s previous estimates were also employed. Hammon's estimates are closer to FAO-PM for December, with Hansen’s following closely. Hargreaves-Samani model, followed by De Bruin model, produced the closest estimates to FAO-PM for August. It is noteworthy that the estimates for August were significantly better than those for December for all empirical models. This is probably because of negligible wind speed values in August. FAO-PM is sensitive in the wind speed values of December. The presented empirical models are quick and accurate and can employ LST inputs of any spatial resolution.

Plants are subjected to a wide range of stresses which reduces the productivity of agricultural crops. In the case of cereal cultivations, climate change impacts on their production mainly through abiotic and biotic stress due for example to heat and water stress but also to pathogens such as bacteria, fungi, nematodes and others. The area under cereal cultivation is increasing worldwide, but, due to these problems, the current rates of yield growth and overall production are not enough to satisfy future demand. For this motivation, there is the needs to monitor and to control the cultivations, also developing new technological solutions useful to better optimize the management strategies, increasing both the quality of products and the quantity of the annual cereal harvest. Infrared imaging is a well-known non-invasive and non-contact technique that represents an outstanding approach of analysis applied in many fields: engineering, medicine, veterinary, cultural heritage and others. In recent years it has been gaining great interest in agriculture as it is well suited to the emerging needs of the precision agriculture management strategies. In this work, we performed an in-field multispectral infrared monitoring of different cereal crops (durum wheat and common wheat) through the use of both LWIR and MWIR cameras. The monitoring carried out made it possible to identify, among the crops analyzed, those subject to higher stress levels and their response to the different spectral ranges used. The results obtained open to the possibility of identifying new figures of merit useful for an effective monitoring of cereal crops and measurable through remote instrumentation.

Many satellite-derived evapotranspiration (ET) estimates rely on coarse resolution (CR) land surface temperature (LST) from 1-km thermal infrared bands offered by NASA and ESA instruments, like Terra MODIS and Sentinel-3 SLSTR. This affects prediction performance of ET, especially in complex regions, such as the Alps. Since most twosource energy balance (TSEB) models assume no thermal variability within a pixel, a major challenge in ET modelling is related to cell grid heterogeneity. Given this limitation, we investigate the potential of kernel-driven downscaling to obtain sub-kilometer LST products based on fine resolution (FR) sensors, i.e., Sentinel-2 MSI and MODIS VNIR, for estimating TSEB-based ET over South Tyrol, in the South-Eastern Alps. To this aim, we exploit relationships between CR LST and FR predictors using trees-based algorithms. Due to reduced capabilities of univariate models in complex ecosystems, multi-source predictors are considered, including multispectral reflectances, spectral indices, solar radiation, and topography. The performance of the TSEB model driven by disaggregated outputs is evaluated against original 1-km LST and ground-based fluxes from two eddy covariance towers. In general, turbulent fluxes forced with downscaled LST resulted in RMSE of 86 Wm-2 and mean bias of 55 Wm-2, which translated to 8% and 15% decrease in the respective estimates when compared to TSEB results with 1-km LST. Despite some limitations, mainly related to small-scale changes in landcover and topography that control LSTs and consequently affect TSEB-based ET estimates, the enhanced land surface temperature has potential for providing energy fluxes at finer spatial resolution in heterogenous ecosystems.