Is deeper always better? Evaluating deep learning models for yield forecasting with small data.

Predicting crop yields, and especially anomalously low yields, is of special importance for food insecure countries. In this study, we investigate a flexible deep learning approach to forecast crop yield at the provincial administrative level based on deep 1D and 2D convolutional neural networks using limited data. This approach meets the operational requirements-public and global records of satellite data in an application ready format with near real time updates-and can be transferred to any country with reliable yield statistics. Three-dimensional histograms of normalized difference vegetation index (NDVI) and climate data are used as input to the 2D model, while simple administrative-level time series averages of NDVI and climate data to the 1D model. The best model architecture is automatically identified during efficient and extensive hyperparameter optimization. To demonstrate the relevance of this approach, we hindcast (2002-2018) the yields of Algeria's three main crops (barley, durum and soft wheat) and contrast the model's performance with machine learning algorithms and conventional benchmark models used in a previous study. Simple benchmarks such as peak NDVI remained challenging to outperform while machine learning models were superior to deep learning models for all forecasting months and all tested crops. We attribute the poor performance of deep learning to the small size of the dataset available.

Comparing land surface phenology of major European crops as derived from SAR and multispectral data of Sentinel-1 and -2.

The frequent acquisitions of fine spatial resolution imagery (10 m) offered by recent multispectral satellite missions, including Sentinel-2, can resolve single agricultural fields and thus provide crop-specific phenology metrics, a crucial information for crop monitoring. However, effective phenology retrieval may still be hampered by significant cloud cover. Synthetic aperture radar (SAR) observations are not restricted by weather conditions, and Sentinel-1 thus ensures more frequent observations of the land surface. However, these data have not been systematically exploited for phenology retrieval so far. In this study, we extracted crop-specific land surface phenology (LSP) from Sentinel-1 and Sentinel-2 of major European crops (common and durum wheat, barley, maize, oats, rape and turnip rape, sugar beet, sunflower, and dry pulses) using ground-truth information from the "Copernicus module" of the Land Use/Cover Area frame statistical Survey (LUCAS) of 2018. We consistently used a single model-fit approach to retrieve LSP metrics on temporal profiles of CR (Cross Ratio, the ratio of the backscattering coefficient VH/VV from Sentinel-1) and NDVI (Normalized Difference Vegetation Index from Sentinel-2). Our analysis revealed that LSP retrievals from Sentinel-1 are comparable to those of Sentinel-2, particularly for winter crops. The start of season (SOS) timings, as derived from Sentinel-1 and -2, are significantly correlated (average r of 0.78 for winter and 0.46 for summer crops). The correlation is lower for end of season retrievals (EOS, r of 0.62 and 0.34). Agreement between LSP derived from Sentinel-1 and -2 varies among crop types, ranging from r = 0.89 and mean absolute error MAE = 10 days (SOS of dry pulses) to r = 0.15 and MAE = 53 days (EOS of sugar beet). Observed deviations revealed that Sentinel-1 and -2 LSP retrievals can be complementary; for example for winter crops we found that SAR detected the start of the spring growth while multispectral data is sensitive to the vegetative growth before and during winter. To test if our results correspond reasonably to in-situ data, we compared average crop-specific LSP for Germany to average phenology from ground phenological observations of 2018 gathered from the German Meteorological Service (DWD). Our study demonstrated that both Sentinel-1 and -2 can provide relevant and at times complementary LSP information at field- and crop-level.

Near real-time vegetation anomaly detection with MODIS NDVI: Timeliness vs. accuracy and effect of anomaly computation options.

For food crises early warning purposes, coarse spatial resolution NDVI data are widely used to monitor vegetation conditions in near real-time (NRT). Different types of NDVI anomalies are typically employed to assess the current state of crops and rangelands as compared to previous years. Timeliness and accuracy of such anomalies are critical factors to an effective monitoring. Temporal smoothing can efficiently reduce noise and cloud contamination in the time series of historical observations, where data points are available before and after each observation to be smoothed. With NRT data, smoothing methods are adapted to cope with the unbalanced availability of data before and after the most recent data points. These NRT approaches provide successive updates of the estimation of the same data point as more observations become available. Anomalies compare the current NDVI value with some statistics (e.g. indicators of central tendency and dispersion) extracted from the historical archive of observations. With multiple updates of the same datasets being available, two options can be selected to compute anomalies, i.e. using the same update level for the NRT data and the statistics or using the most reliable update for the latter. In this study we assess the accuracy of three commonly employed 1 km MODIS NDVI anomalies (standard scores, non-exceedance probability and vegetation condition index) with respect to (1) delay with which they become available and (2) option selected for their computation. We show that a large estimation error affects the earlier estimates and that this error is efficiently reduced in subsequent updates. In addition, with regards to the preferable option to compute anomalies, we empirically observe that it depends on the type of application (e.g. averaging anomalies value over an area of interest vs. detecting "drought" conditions by setting a threshold on the anomaly value) and the employed anomaly type. Finally, we map the spatial pattern in the magnitude of NRT anomaly estimation errors over the globe and relate it to average cloudiness.

Remote sensing of solar-induced chlorophyll fluorescence (SIF) in vegetation: 50 years of progress.

Remote sensing of solar-induced chlorophyll fluorescence (SIF) is a rapidly advancing front in terrestrial vegetation science, with emerging capability in space-based methodologies and diverse application prospects. Although remote sensing of SIF - especially from space - is seen as a contemporary new specialty for terrestrial plants, it is founded upon a multi-decadal history of research, applications, and sensor developments in active and passive sensing of chlorophyll fluorescence. Current technical capabilities allow SIF to be measured across a range of biological, spatial, and temporal scales. As an optical signal, SIF may be assessed remotely using highly-resolved spectral sensors and state-of-the-art algorithms to distinguish the emission from reflected and/or scattered ambient light. Because the red to far-red SIF emission is detectable non-invasively, it may be sampled repeatedly to acquire spatio-temporally explicit information about photosynthetic light responses and steady-state behaviour in vegetation. Progress in this field is accelerating with innovative sensor developments, retrieval methods, and modelling advances. This review distills the historical and current developments spanning the last several decades. It highlights SIF heritage and complementarity within the broader field of fluorescence science, the maturation of physiological and radiative transfer modelling, SIF signal retrieval strategies, techniques for field and airborne sensing, advances in satellite-based systems, and applications of these capabilities in evaluation of photosynthesis and stress effects. Progress, challenges, and future directions are considered for this unique avenue of remote sensing.

ASAP: A new global early warning system to detect anomaly hot spots of agricultural production for food security analysis.

Monitoring crop and rangeland conditions is highly relevant for early warning and response planning in food insecure areas of the world. Satellite remote sensing can obtain relevant and timely information in such areas where ground data are scattered, non-homogenous, or frequently unavailable. Rainfall estimates provide an outlook of the drivers of vegetation growth, whereas time series of satellite-based biophysical indicators at high temporal resolution provide key information about vegetation status in near real-time and over large areas. The new early warning decision support system ASAP (Anomaly hot Spots of Agricultural Production) builds on the experience of the MARS crop monitoring activities for food insecure areas, that have started in the early 2000's and aims at providing timely information about possible crop production anomalies. The information made available on the website (https://mars.jrc.ec.europa.eu/asap/) directly supports multi-agency early warning initiatives such as for example the GEOGLAM Crop Monitor for Early Warning and provides inputs to more detailed food security assessments that are the basis for the annual Global Report on Food Crises. ASAP is a two-step analysis framework, with a first fully automated step classifying the first sub-national level administrative units into four agricultural production deficit warning categories. Warnings are based on rainfall and vegetation index anomalies computed over crop and rangeland areas and are updated every 10 days. They take into account the timing during the crop season at which they occur, using remote sensing derived phenology per-pixel. The second step involves the monthly analysis at country level by JRC crop monitoring experts of all the information available, including the automatic warnings, crop production and food security-tailored media analysis, high-resolution imagery (e.g. Landsat 8, Sentinel 1 and 2) processed in Google Earth Engine and ancillary maps, graphs and statistics derived from a set of indicators. Countries with potentially critical conditions are marked as minor or major hotspots and a global overview is provided together with short national level narratives.

Remote sensing monitoring of land restoration interventions in semi-arid environments with a before-after control-impact statistical design.

Restoration interventions to combat land degradation are carried out in arid and semi-arid areas to improve vegetation cover and land productivity. Evaluating the success of an intervention over time is challenging due to various constraints (e.g. difficult-to-access areas, lack of long-term records) and the lack of standardised and affordable methodologies. We propose a semi-automatic methodology that uses remote sensing data to provide a rapid, standardised and objective assessment of the biophysical impact, in terms of vegetation cover, of restoration interventions. The Normalised Difference Vegetation Index (NDVI) is used as a proxy for vegetation cover. Recognising that changes in vegetation cover are naturally due to environmental factors such as seasonality and inter-annual climate variability, conclusions about the success of the intervention cannot be drawn by focussing on the intervention area only. We therefore use a comparative method that analyses the temporal variations (before and after the intervention) of the NDVI of the intervention area with respect to multiple control sites that are automatically and randomly selected from a set of candidates that are similar to the intervention area. Similarity is defined in terms of class composition as derived from an ISODATA classification of the imagery before the intervention. The method provides an estimate of the magnitude and significance of the difference in greenness change between the intervention area and control areas. As a case study, the methodology is applied to 15 restoration interventions carried out in Senegal. The impact of the interventions is analysed using 250-m MODIS and 30-m Landsat data. Results show that a significant improvement in vegetation cover was detectable only in one third of the analysed interventions, which is consistent with independent qualitative assessments based on field observations and visual analysis of high resolution imagery. Rural development agencies may potentially use the proposed method for a first screening of restoration interventions.

Ground-based optical measurements at European flux sites: a review of methods, instruments and current controversies.

This paper reviews the currently available optical sensors, their limitations and opportunities for deployment at Eddy Covariance (EC) sites in Europe. This review is based on the results obtained from an online survey designed and disseminated by the Co-cooperation in Science and Technology (COST) Action ESO903-"Spectral Sampling Tools for Vegetation Biophysical Parameters and Flux Measurements in Europe" that provided a complete view on spectral sampling activities carried out within the different research teams in European countries. The results have highlighted that a wide variety of optical sensors are in use at flux sites across Europe, and responses further demonstrated that users were not always fully aware of the key issues underpinning repeatability and the reproducibility of their spectral measurements. The key findings of this survey point towards the need for greater awareness of the need for standardisation and development of a common protocol of optical sampling at the European EC sites.

Characterization of fine resolution field spectrometers using solar Fraunhofer lines and atmospheric absorption features.

The accurate spectral characterization of high-resolution spectrometers is required for correctly computing, interpreting, and comparing radiance and reflectance spectra acquired at different times or by different instruments. In this paper, we describe an algorithm for the spectral characterization of field spectrometer data using sharp atmospheric or solar absorption features present in the measured data. The algorithm retrieves systematic shifts in channel position and actual full width at half-maximum (FWHM) of the instrument by comparing data acquired during standard field spectroscopy measurement operations with a reference irradiance spectrum modeled with the MODTRAN4 radiative transfer code. Measurements from four different field spectrometers with spectral resolutions ranging from 0.05 to 3.5nm are processed and the results validated against laboratory calibration. An accurate retrieval of channel position and FWHM has been achieved, with an average error smaller than the instrument spectral sampling interval.

Quantifying and addressing the prevalence and bias of study designs in the environmental and social sciences.

Building trust in science and evidence-based decision-making depends heavily on the credibility of studies and their findings. Researchers employ many different study designs that vary in their risk of bias to evaluate the true effect of interventions or impacts. Here, we empirically quantify, on a large scale, the prevalence of different study designs and the magnitude of bias in their estimates. Randomised designs and controlled observational designs with pre-intervention sampling were used by just 23% of intervention studies in biodiversity conservation, and 36% of intervention studies in social science. We demonstrate, through pairwise within-study comparisons across 49 environmental datasets, that these types of designs usually give less biased estimates than simpler observational designs. We propose a model-based approach to combine study estimates that may suffer from different levels of study design bias, discuss the implications for evidence synthesis, and how to facilitate the use of more credible study designs.

Modeling gross primary production of agro-forestry ecosystems by assimilation of satellite-derived information in a process-based model.

In this paper we present results obtained in the framework of a regional-scale analysis of the carbon budget of poplar plantations in Northern Italy. We explored the ability of the process-based model BIOME-BGC to estimate the gross primary production (GPP) using an inverse modeling approach exploiting eddy covariance and satellite data. We firstly present a version of BIOME-BGC coupled with the radiative transfer models PROSPECT and SAILH (named PROSAILH-BGC) with the aims of i) improving the BIOME-BGC description of the radiative transfer regime within the canopy and ii) allowing the assimilation of remotely-sensed vegetation index time series, such as MODIS NDVI, into the model. Secondly, we present a two-step model inversion for optimization of model parameters. In the first step, some key ecophysiological parameters were optimized against data collected by an eddy covariance flux tower. In the second step, important information about phenological dates and about standing biomass were optimized against MODIS NDVI. Results obtained showed that the PROSAILH-BGC allowed simulation of MODIS NDVI with good accuracy and that we described better the canopy radiation regime. The inverse modeling approach was demonstrated to be useful for the optimization of ecophysiological model parameters, phenological dates and parameters related to the standing biomass, allowing good accuracy of daily and annual GPP predictions. In summary, this study showed that assimilation of eddy covariance and remote sensing data in a process model may provide important information for modeling gross primary production at regional scale.

Assessing Steady-state Fluorescence and PRI from Hyperspectral Proximal Sensing as Early Indicators of Plant Stress: The Case of Ozone Exposure.

High spectral resolution spectrometers were used to detect optical signals ofongoing plant stress in potted white clover canopies subjected to ozone fumigation. Thecase of ozone stress is used in this manuscript as a paradigm of oxidative stress. Steadystatefluorescence (Fs) and the Photochemical Reflectance Index (PRI) were investigatedas advanced hyperspectral remote sensing techniques able to sense variations in the excessenergy dissipation pathways occurring when photosynthesis declines in plants exposed to astress agent. Fs and PRI were monitored in control and ozone fumigated canopies during a21-day experiment together with the traditional Normalized Difference Vegetation Index(NDVI) and physiological measurements commonly employed by physiologists to describestress development (i.e. net CO2 assimilation, active fluorimetry, chlorophyll concentrationand visible injuries). It is shown that remote detection of an ongoing stress through Fs andPRI can be achieved in an early phase, characterized by the decline of photosynthesis. Onthe contrary, NDVI was able to detect the stress only when damage occurred. These resultsopen up new possibilities for assessment of plant stress by means of hyperspectral remotesensing.

Assessment of oak forest condition based on leaf biochemical variables and chlorophyll fluorescence.

Pedunculate oak forests (Quercus robur L.) in the Ticino Regional Park, Italy, are declining as a result of insect attacks, summer droughts and air pollution. The assessment and monitoring of forest condition can provide a basis for managing and conserving forest ecosystems and thereby avoid loss of valuable natural resources. Currently, most forest assessments are limited to ground-based visual evaluations that are local and subjective. It is therefore difficult to compare data collected by different crews or to define reliable trends over years. We examined vegetation variables that can be quantitatively estimated by remote observations and, thus, are suitable for objective monitoring over extended forested areas. We found that total chlorophyll (Chl) concentration is the most suitable variable for assessing pedunculate oak decline. It is highly correlated with visual assessments of discoloration. Furthermore, Chl concentration can be accurately estimated from leaf optical properties, making it feasible to map Chl concentration at the canopy level from satellite and airborne remote observations.

Revisiting the concept of a symmetric index of agreement for continuous datasets.

Quantifying how close two datasets are to each other is a common and necessary undertaking in scientific research. The Pearson product-moment correlation coefficient r is a widely used measure of the degree of linear dependence between two data series, but it gives no indication of how similar the values of these series are in magnitude. Although a number of indexes have been proposed to compare a dataset with a reference, only few are available to compare two datasets of equivalent (or unknown) reliability. After a brief review and numerical tests of the metrics designed to accomplish this task, this paper shows how an index proposed by Mielke can, with a minor modification, satisfy a series of desired properties, namely to be adimensional, bounded, symmetric, easy to compute and directly interpretable with respect to r. We thus show that this index can be considered as a natural extension to r that downregulates the value of r according to the bias between analysed datasets. The paper also proposes an effective way to disentangle the systematic and the unsystematic contribution to this agreement based on eigen decompositions. The use and value of the index is also illustrated on synthetic and real datasets.

Using optical remote sensing techniques to track the development of ozone-induced stress.

In this paper, a literature review about optical remote sensing (RS) of O(3) stress is presented. Studies on O(3)-induced effects on vegetation reflectance have been conducted since late '70s based on the analysis of optical RS data. Literature review reveals that traditional RS techniques were able to detect changes in leaf and canopy reflectance related to O(3)-induced stress when visible symptoms already occurred. Only recently, advanced RS techniques using hyperspectral sensors, demonstrated the feasibility of detecting the stress in its early phase by monitoring excess energy dissipation pathways such as chlorophyll fluorescence and non-photochemical quenching (NPQ). Steady-state fluorescence (Fs), measured by exploiting the Fraunhofer line depth principle and NPQ related xanthophyll-cycle, estimated through the photochemical reflectance index (PRI) responded to O(3) fumigation before visible symptoms occurred. This opens up new possibilities for the early detection of vegetation O(3) stress by means of hyperspectral RS.

A flux-based assessment of the effects of ozone on foliar injury, photosynthesis, and yield of bean (Phaseolus vulgaris L. cv. Borlotto Nano Lingua di Fuoco) in open-top chambers.

Stomatal ozone uptake, determined with the Jarvis' approach, was related to photosynthetic efficiency assessed by chlorophyll fluorescence and reflectance measurements in open-top chamber experiments on Phaseolus vulgaris. The effects of O(3) exposure were also evaluated in terms of visible and microscopical leaf injury and plant productivity. Results showed that microscopical leaf symptoms, assessed as cell death and H(2)O(2) accumulation, preceded by 3-4 days the appearance of visible symptoms. An effective dose of ozone stomatal flux for visible leaf damages was found around 1.33 mmol O(3) m(-2). Significant linear dose-response relationships were obtained between accumulated fluxes and optical indices (PRI, NDI, DeltaF/F'(m)). The negative effects on photosynthesis reduced plant productivity, affecting the number of pods and seeds, but not seed weight. These results, besides contributing to the development of a flux-based ozone risk assessment for crops in Europe, highlight the potentiality of reflectance measurements for the early detection of ozone stress.