Search

Agriculture monitoring and yield estimation are important for food security, both regionally and globally.Understanding the year-to-year variability in crop yield and its relationship to meteorological conditions isparticularly important for several regions where yield is highly dependent on changing environmental factors.For example, wheat yields in India have been steadily increasing since the 1960s and 1970s due to benefits ofthe Green Revolution, but in recent years the wheat yield has been unstable as major crop yield losses wereattributed to unfavorable meteorological conditions. Modeling the effects of various environmental variables canbe challenging, as their impact on the final yield is complex and varies depending on their intensity and thecrop growth stage at which they occur (e.g., moderate rainfall is beneficial for crops, but extensive and untimelyrainfalls can lead to huge yield losses).In this work, we exploit interconnections between meteorological conditions and satellite data on vegeta-tion during the whole growing season, and their simultaneous impact on wheat yield in the Wheat Belt inIndia. We use GLDAS 2.1 data as the meteorological input and MODIS data for the vegetation remote sensinginput. Adding satellite information on crop is crucial for yield estimation, as it carries information on both cropphenology, as well as the crop response to the meteorological conditions. We apply machine learning algorithms(e.g., convolutional neural networks, CNNs) that can model non-linear processes and can extract importantfeatures in the multivariate time series automatically from data, without prior knowledge or human effort in featuredesign. By doing so, we do not force assumptions on which time is the most important for the final crop yieldand we can include in the analysis the whole time series of multiple input variables at a high temporal resolution.Furthermore, we analyze the CNNs in terms of important features and crucial time windows for yield estimation,which shows that they the vary across space and time. By combining meteorological and satellite vegetation datawith CNNs this work may help to disentangle the complex interactions between the features in the time series ofthe input data and the wheat yield.

Agriculture monitoring and yield estimation are important for food security, both globally and locally. Understanding year-to-year variability in crop yield and it’s relationship to meteorological conditions is especially important for countries and regions where the yield losses due to environmental factors are already observed and further predicted under the impact of climate change. One of such countries is India, where wheat is a staple food for a large part of the population and where the wheat production is the second largest in the world. Wheat yields in India have been steadily increasing since 1960s and 1970s due to benefits of the Green Revolution. However, in recent years the wheat yield has been unstable as major crop yield losses were attributed to unfavorable meteorological conditions. Including various environmental variables in the crop yield modeling can be challenging, as their impact on the final yield is complex and varies depending on the crop growth stage. For example, moderate rainfall is beneficial for crops, but extensive and untimely rainfalls can lead to huge yield losses, as in 2015, which points out the importance of accounting for such non-linear processes. In this work, we exploit the relationship of meteorological conditions during the whole growing season simultaneously with the satellite data on vegetation, and the wheat yield in the Wheat Belt in India. We use GLDAS 2.1 data as the meteorological input and MODIS data for the vegetation remote sensing input. We show that adding satellite information on crop is crucial for yield estimation, as it carries information on both crop phenology, as well as on the crop response to the meteorological conditions. We apply the convolutional neural networks (CNNs) that can model non-linear processes and can extract important features on their own, without prior knowledge or human effort in feature design. This also means that we do not force assumptions on which time is the most important for the final crop yield. By doing so, we can include in the analysis the whole time series of multiple input variables at a high temporal resolution (4 days), and model nonlinear interactions among them. This is especially important for the studied region, as the timing of wheat planting and ripening vary across the Indian counties and inter-annually due to farmer practices and various meteorological conditions. Furthermore, we analyze the CNNs in terms of important features and important time windows for yield estimation, which shows that these vary across space and time. By combining meteorological and satellite vegetation data with CNNs this work may help to disentangle the complex interactions between the features in the time series of the input data and the wheat yield.

The SPARC campaign has been organized in coincidence of CHRIS/Proba multi-angular and hyperspectral acquisition over the agricultural test site of Barrax in Spain. Radiometric and biophysical vegetation parameters measurements have been carried out for different crops in both field and laboratory. . The aim of this preliminary study is to assess the capability to estimate Leaf Area Index (LAI) from CHRIS/Proba data by inverting the coupled radiative transfer models PROSPECT+SAILH. Two different approaches have been used to invert canopy reflectance observed in 5 directions and 62 spectral bands: Look Up Tables and Pest ASP Tool. Results show that the use of a priori knowledge is of great importance for the estimation of LAI. In this way we have been able to estimate this parameter with an accuracy of around 15%+20% for Alfalfa, Potatoes, and Sugar Beet samples.

Hyperspectral images are of increasing importance in remote sensing applications. Imaging spectrometers provide semi-continuous spectra that can be used for physics based surface cover material identification and quantification. Preceding radiometric calibrations serve as a basis for the transformation of measured signals into physics based units such as radiance. Pushbroom sensors collect incident radiation by at least one detector array utilizing the photoelectric effect. Temporal variations of the detector characteristics that differ with foregoing radiometric calibration cause visually perceptible along-track stripes in the at-sensor radiance data that aggravate succeeding image-based analyses. Especially, variations of the thermally induced dark current dominate and have to be reduced. In this work, a new approach is presented that efficiently reduces dark current related stripe noise. It integrates an across-effect gradient minimization principle. The performance has been evaluated using artificially degraded whiskbroom (reference) and real pushbroom acquisitions from EO-1 Hyperion and AISA DUAL that are significantly covered by stripe noise. A set of quality indicators has been used for the accuracy assessment. They clearly show that the new approach outperforms a limited set of tested state-of-the-art approaches and achieves a very high accuracy related to ground-truth for selected tests. It may substitute recent algorithms in the Reduction of Miscalibration Effects (ROME) framework that is broadly used to reduce radiometric miscalibrations of pushbroom data takes.

New recently launched and upcoming satellite missions offer new possibilities to monitor vegetation from space. Among them, sun-induced fluorescence (SIF) of chlorophyll-a has provided a novel way to track plant photosynthesis from space. Unfortunately, for now, SIF has been only observed with instruments that provide data at high spectral resolution, but at the cost of low spatial resolution of tens of kilometers. However, soon applicability of SIF to observe vegetation will increase with the launch of Tropomi on board Sentinel-5 Precursor that will provide data at a much improved spatio-temporal scale, and FLuorescence EXplorer (FLEX), the first mission ever specifically designed to retrieve SIF, planned to be launched in 2022. SIF data have been shown to be sensitive to photosynthetic activities, and for croplands even more than other vegetation. Furthermore, SIF data have used to infer crop productivity directly, or to improve estimates at the flux towers, by using them together with retrieved vegetation indices and on site measurements of other vegetation properties. Here, we will show results of integrating SIF data at low spatial resolution with the reflectance data from Sentinel 2 at an excellent spatial resolution of tens of meters. The leaf and canopy parameters were derived from Sentinel 2 reflectance data with a look-up-table approach, where the combined PROSPECT leaf optical properties model and 4SAIL canopy bidirectional reflectance model were used for the inversion. The retrieved parameters are subsequently used together with space-borne SIF within the Soil-Canopy Observation of Photosynthesis and Energy (SCOPE) balance model to derive photosynthetic traits and primary production. We will present our first results of the ongoing work, for selected crops regions in the Corn Belt in North America and in Europe.

© The Author(s) 2016. Each year, terrestrial ecosystems absorb more than a quarter of the anthropogenic carbon emissions, termed as land carbon sink. An exceptionally large land carbon sink anomaly was recorded in 2011, of which more than half was attributed to Australia. However, the persistence and spatially attribution of this carbon sink remain largely unknown. Here we conducted an observation-based study to characterize the Australian land carbon sink through the novel coupling of satellite retrievals of atmospheric CO2 and photosynthesis and in-situ flux tower measures. We show the 2010-11 carbon sink was primarily ascribed to savannas and grasslands. When all biomes were normalized by rainfall, shrublands however, were most efficient in absorbing carbon. We found the 2010-11 net CO2 uptake was highly transient with rapid dissipation through drought. The size of the 2010-11 carbon sink over Australia (0.97 Pg) was reduced to 0.48 Pg in 2011-12, and was nearly eliminated in 2012-13 (0.08 Pg). We further report evidence of an earlier 2000-01 large net CO2 uptake, demonstrating a repetitive nature of this land carbon sink. Given a significant increasing trend in extreme wet year precipitation over Australia, we suggest that carbon sink episodes will exert greater future impacts on global carbon cycle.

Classification of clouds, cirrus, snow, shadows and clear sky areas is a crucial step in the pre-processing of optical remote sensing images and is a valuable input for their atmospheric correction. The Multi-Spectral Imager on board the Sentinel-2’s of the Copernicus program offers optimized bands for this task and delivers unprecedented amounts of data regarding spatial sampling, global coverage, spectral coverage, and repetition rate. Efficient algorithms are needed to process, or possibly reprocess, those big amounts of data. Techniques based on top-of-atmosphere reflectance spectra for single-pixels without exploitation of external data or spatial context offer the largest potential for parallel data processing and highly optimized processing throughput. Such algorithms can be seen as a baseline for possible trade-offs in processing performance when the application of more sophisticated methods is discussed. We present several ready-to-use classification algorithms which are all based on a publicly available database of manually classified Sentinel-2A images. These algorithms are based on commonly used and newly developed machine learning techniques which drastically reduce the amount of time needed to update the algorithms when new images are added to the database. Several ready-to-use decision trees are presented which allow to correctly label about 91 % of the spectra within a validation dataset. While decision trees are simple to implement and easy to understand, they offer only limited classification skill. It improves to 98 % when the presented algorithm based on the classical Bayesian method is applied. This method has only recently been used for this task and shows excellent performance concerning classification skill and processing performance. A comparison of the presented algorithms with other commonly used techniques such as random forests, stochastic gradient descent, or support vector machines is also given. Especially random forests and support vector machines show similar classification skill as the classical Bayesian method.

This work presents a new methodology that faces the problem of accurate identification of location and abundance of clouds in multispectral images acquired by space-borne sensors working in the visible and near-infrared (VNIR) spectral range. The amount of images acquired over the globe every day by the instruments on board Earth Observation satellites makes inevitable that many of these images present cloud covers. The objective of this work is to develop and validate a method that takes advantage of the high spectral and radiometric resolution, and the specific band locations (e.g. the oxygen band) of present multispectral sensors to increase the cloud detection accuracy. Moreover, the method provides probability and cloud abundance rather than flags, which can be used to describe detected clouds (subpixel coverage, cloud type, height, etc) more accurately.

A characteristic spectral emission is observed in vegitation chlorophyll under excitation by solar radiation. This emission, known as solar-induced chlorophyll fluorescence, occurs in the red and near infra-red spectral regions. In this paper a new methodology for the estimation of solar-induced chlorophyll fluorescence from spaceborne and airborne sensors is presented. The fluorescence signal is included in an atmospheric radiative transfer scheme so that chlorophyll fluorescence and surface reflectance are retrieved consistently from the measured at-sensor radiance. This methodology is tested on images acquired by the Medium Resolution Imaging Spectrometer (MERIS) on board the ENVIronmental SATellite (ENVISAT) taking advantage of its good charactefization of the O2-A absorption band. Validation of MERIS-derived fluorescence is carried out by applying the method to data acquired by the Compact Airborne Spectrographic Imager (CASI-1500) sensor concurrently to MERIS acquisitions. CASI-derived fluorescence is in turn compared with ground-based fluorescence measurements, a correlation coefficient R2 of 0.85 being obtained. Copyright 2007 by the American Geophysical Union.

The SCAPE-M method has been developed for the retrieval of aerosol optical thickness (AOT), columnar water vapor (CWV) and surface reflectance from MERIS data. The retrieval of the three products is performed on a coupled manner taking profit of a common basis in which variations in surface elevation, topography and atmospheric state are considered pixel-by-pixel. More than 200 MERIS images have been processed by SCAPE-M. This results in a large data base of AOT, CWV and reflectance data to be validated and analysed. Comparison of SCAPE-M AOT retrievals with data from AErosol RObotic NETwork (AERONET) stations showed typical Root Mean Square Errors (RMSEs) of 0.05, an average correlation coefficient R2 of about 0.7-0.8. The R? values grow up to more than 0.9 in the case of CWV after comparison with the same stations. Retrieved surface reflectance maps have been successfully compared with reflectance data derived from CHRIS-PROBA images and with ground-based measurements over water.

The CEFLES2 campaign during the Carbo Europe Regional Experiment Strategy was designed to provide simultaneous airborne measurements of solar induced fluorescence and CO2 fluxes. It was combined with extensive ground-based quantification of leaf- and canopy-level processes in support of ESA's Candidate Earth Explorer Mission of the "Fluorescence Explorer" (FLEX). The aim of this campaign was to test if fluorescence signal detected from an airborne platform can be used to improve estimates of plant mediated exchange on the mesoscale. Canopy fluorescence was quantified from four airborne platforms using a combination of novel sensors: (i) the prototype airborne sensor AirFLEX quantified fluorescence in the oxygen A and B bands, (ii) a hyperspectral spectrometer (ASD) measured reflectance along transects during 12 day courses, (iii) spatially high resolution georeferenced hyperspectral data cubes containing the whole optical spectrum and the thermal region were gathered with an AHS sensor, and (iv) the first employment of the high performance imaging spectrometer HYPER delivered spatially explicit and multi-temporal transects across the whole region. During three measurement periods in April, June and September 2007 structural, functional and radiometric characteristics of more than 20 different vegetation types in the Les Landes region, Southwest France, were extensively characterized on the ground. The campaign concept focussed especially on quantifying plant mediated exchange processes (photosynthetic electron transport, CO2 uptake, evapotranspiration) and fluorescence emission. The comparison between passive sun-induced fluorescence and active laser-induced fluorescence was performed on a corn canopy in the daily cycle and under desiccation stress. Both techniques show good agreement in detecting stress induced fluorescence change at the 760 nm band. On the large scale, airborne and ground-level measurements of fluorescence were compared on several vegetation types supporting the scaling of this novel remote sensing signal. The multi-scale design of the four airborne radiometric measurements along with extensive ground activities fosters a nested approach to quantify photosynthetic efficiency and gross primary productivity (GPP) from passive fluorescence.

In the framework of preparatory activities for the SPECTRA (Surface Processes and Ecosystems Changes Through Response Analysis) ESA Earth Explorer Core Mission, CHRIS/PROBA acquisitions over the Barrax Core Site in Spain were used to compile a reference dataset for future in-depth studies. Taking advantage of the possibility of consecutive days of acquisitions, multiple-angular acquisitions finally included 10 different view angles from CHRIS, in Mode 1 with 62 spectral, and a ground resolution of about 34 m. Additional ROSIS and HYMAP sensors, flying simultaneously with CHRIS overpass, provided detailed images for validation of CHRIS data, particularly in the spectral domain. Moreover, up to 3 angles per sample from airborne HYMAP data were acquired, with high spectral and spatial resolution, and then both spectral and angular domains can be exploited with the combined CHRIS/HYMAP/ROSIS dataset. Detailed soil/vegetation and atmospheric measurements complete the SPARC data, and data from other satellites (MERIS, SEVIRI, SPOT, Landsat) were collected as well, to address scaling issues. Methods for data analysis and exploitation have been developed in the context of SPARC activities, and preliminary results about retrievals of biophysical information from multi-angular hyperspectral data are already available. The whole SPARC dataset represents a reference for the exploitation of CHRIS data, allowing the development of new processing and retrieval algorithms, and the validation of such algorithms by means of ground measurements and complementary airborne and satellite data. More details on several processing aspects of the CHRIS/PROBA data acquired within the SPARC campaign are presented in other papers in this conference.

With the advent of dedicated greenhouse-gas space-borne spectrometers sporting high resolution spectra in the O₂ A-band spectral region (755-774 nm), the retrieval of chlorophyll fluorescence has become feasible on a global scale. If unaccounted for, however, fluorescence can indirectly perturb the greenhouse gas retrievals as it perturbs the oxygen absorption features. As atmospheric CO₂ measurements are used to invert net fluxes at the land-atmosphere interface, a bias caused by fluorescence can be crucial as it will spatially correlate with the fluxes to be inverted. Avoiding a bias and retrieving fluorescence accurately will provide additional constraints on both the net and gross fluxes in the global carbon cycle. We show that chlorophyll fluorescence, if neglected, systematically interferes with full-physics multi-band XCO₂ retrievals using the O₂ A-band. Systematic biases in XCO₂ can amount to + 1 ppm if fluorescence constitutes 1% to the continuum level radiance. We show that this bias can be largely eliminated by simultaneously fitting fluorescence in a full-physics based retrieval.If fluorescence is the primary target, a dedicated but very simple retrieval based purely on Fraunhofer lines is shown to be more accurate and very robust even in the presence of large scattering optical depths. We find that about 80% of the surface fluorescence is retained at the top-of-atmosphere even for cloud optical thicknesses around 2-5. We further show that small instrument modifications to future O₂ A-band spectrometer spectral ranges can result in largely reduced random errors in chlorophyll fluorescence, paving the way towards a more dedicated instrument exploiting solar absorption features only.

The scope of the Science Plan is to describe the scientific background, applications, and activities related to the Environmental Mapping and Analysis Program (EnMAP) mission. Primarily, the document addresses scientists and funding institutions, but it may also be of interest for environmental stakeholders and governmental bodies. It is conceived to be a living document that will be updated throughout the entire mission. Chapter 1 provides a brief overview of the principles and current state of imaging spectroscopy. This is followed by an introduction to the EnMAP mission, including its objectives and potential impact on international programs as well as major environmental and societal challenges to their understanding and management EnMAP can contribute. Chapter 2 describes the EnMAP system together with data products and access, calibration/validation issues, and synergies with other missions. Chapter 3 gives an overview of the relevance, current lines of research, and potential contributions of EnMAP for major fields of application, such as vegetation, geology and soils, coastal and inland waters, cryosphere, urban areas, atmosphere and hazards to address the environmental and societal challenges presented in Chapter 1. Finally, Chapter 4 outlines the scientific exploitation strategy, which includes the strategy for community building and training, preparatory flight campaigns and software developments. A list of abbreviations is provided in the annex to this document, while an extended glossary of terms and abbreviations is available at the EnMAP website.

The consequences of climate change represent a major threat to sustainable development and growth in Southern Africa. Understanding the impact on the geo- and biosphere is therefore of great importance in this particular region. In this context the Kalahari salt pans (also known as playas or sabkhas) and their peripheral saline and alkaline habitats are an ecosystem of major interest. They are very sensitive to environmental conditions, and as thus hydrological, mineralogical and ecological responses to climatic variations can be analysed. Up to now the soil composition of salt pans in this area have been only assessed mono-temporally and on a coarse regional scale. Furthermore, the dynamic of the salt pans, especially the formation of evaporites, is still uncertain and poorly understood. High spectral resolution remote sensing can estimate evaporite content and mineralogy of soils based on the analyses of the surface reflectance properties within the Visible-Near InfraRed (VNIR 400-1000 nm) and Short-Wave InfraRed (SWIR 1000-2500 nm) regions. In these wavelength regions major chemical components of the soil interact with the electromagnetic radiation and produce characteristic absorption features that can be used to derive the properties of interest. Although such techniques are well established for the laboratory and field scale, the potential of current (Hyperion) and upcoming spaceborne sensors such as EnMAP for quantitative mineralogical and salt spectral mapping is still to be demonstrated. Combined with hyperspectral methods, multitemporal remote sensing techniques allow us to derive the recent dynamic of these salt pans and link the mineralogical analysis of the pan surface to major physical processes in these dryland environments. In this study we focus on the analyses of the Namibian Omongwa salt pans based on satellite hyperspectral imagery and multispectral time-series data. First, a change detection analysis is applied using the Iterative-reweighted Multivariate Alteration Detection (iMAD) method to identify and investigate surface changes based on Landsat archive imagery covering the period 1984-2015. For the complete Landsat time series, a total of 130 bi-temporal change maps have been derived and are compiled to produce a change magnitude map, which indicates different regions of activity and stability of the pan surface. Then radiometric geometric and atmospheric correction of Hyperion data were performed using the pre-processing chain of Rogass et al. (2014). Based on field spectroscopy and XRD analysis of soil samples, one of the main evaporite minerals of the pan soil could be identified as gypsum. Different approaches based on spectral features were tested and validated against reference samples to find the most suitable approach for estimating the gypsum content from the Hyperion data. Major challenges were the influence of water vapour absorption bands close to major gypsum absorption feature as well as low signal-to-noise ratio of Hyperion in the SWIR wavelength. A best method was determined that allow to determine Gypsum content with R² of 0.7 and relative RMSE of 0.14. Results reveal a variable spatial distribution of different mineralogy and in particular gypsum content within the pan, which seems to be associated with varying dynamic of the pan surface.

We describe EnMAP-like imaging spectroscopy data files to be used for mineral mapping with the EnMAPBOX software. It is simulated EnMAP satellite data, which is based on hyperspectral flight cam-paign data with the AVIRIS-NG and HyMap sensors. In preparation of the EnMAP satellite mission, an EnMAPBOX software package provides tools for visualization and scientific analysis of the data. Among many applications, the EnMAPBOX contains geological mapping tools (EnGeoMAP). Here we apply these tools to several representative test cases (Boesche, 2015; Boesche et al., 2016; Mielke et al., 2016). The test data comprise two study sites. The first scene covers the Mountain Pass open pit mine - a carbonatite deposit in California, USA. It contains calcitic rock units and rare earth element (REE) bearing minerals of the bastnaesite group, also called fluorocarbonates (Olson et al., 1954). The REE concentrations at mountain pass are 9.2% on average, among the highest in the world (Brüning and Böhmer, 2011). The high concentration and the open pit activities make Mountain Pass an ideal test site to investigate the rare earth element distribution in the surface layer. The airborne image data were collected in 2014 by Jet Propulsion Laboratory (JPL), USA, with the AVIRIS-NG sensor and form the basis for EnMAP simulations (Segl et al., 2012; Thompson et al., 2015). The second HyMap spectral image data covers part of the Miocene Cabo de Gata-Nίjar volcanic field, in southeast Spain. It comprises a subset of (Chabrillat et al., 2016) covering the Rodalquilar and Lomilla Calderas, which host the economically relevant gold-silver, lead-zinc-silver-gold and alunite deposits. It is a hydrothermal alteration complex, representing the silicic alteration, the advanced argillic alter-ation zone, which grades into the argillic and propylitic zone (Arribas et al., 1995, 1989). The image data are part of the Cabo de Gata-Nίjar HyMap imagery which was collected during the DLR HyEurope airborne campaign 2005 in the frame of the GFZ land degradation program (Chabrillat et al., 2016, 2005). We use these datasets to simulate EnMAP-like images for classification and mapping using spectro-scopic remote sensing techniques in the EnGeoMAP tools. The EnMAP end-to-end Simulation (EeteS) tool produced simulated EnMAP like data with a spatial sampling distance of 30 x 30 m and 242 spectral bands (Guanter et al., 2015; Segl et al., 2012).