Changes in seasonal timing affect species and ecosystem response to environmental change. Observations of plant and animal phenology as well as remote sensing and modeling studies document complex interactions and raise many open questions.
We invite contributions with cross-disciplinary perspectives that address seasonality changes based on recent plant and animal phenological observations, pollen monitoring, historical documentary sources, or seasonality measurements using climate data, remote sensing, flux measurements or modeling studies. Contributions across all spatial and temporal scales are welcome that compare and integrate seasonality changes, study effects of long-term climate change or single extreme events, emphasize applications and phenology informed decision-making, discuss species interactions and decoupling, advance our understanding of how seasonality change affects carbon budgets and atmosphere/biosphere feedbacks, and integrate phenology into Earth System Models.
We emphasize phenology informed applications for decision-making and environmental assessment, public health, agriculture and forest management, mechanistic understanding of the phenological processes, and effects of changing phenology on biomass production and carbon budgets. We also welcome contributions addressing international collaboration and program-building initiatives including citizen science networks and data analyses.
This session is organized by a consortium representing the International Society of Biometeorology (Phenology Commission), the Pan-European Phenology Network - PEP 725, the Swiss Academy of Science SCNAT, the TEMPO French Phenology Network and the USA National Phenology Network.
vPICO presentations: Wed, 28 Apr
Phenological shifts in plants greatly affect biotic interactions and lead to multiple feedbacks to the climate system. Increases in growing-season length under warmer climates are expected to drive changes in water, nutrient, and energy fluxes as well as enhancing ecosystem carbon uptake. Yet, future trajectories of growing-season lengths remain highly uncertain because the intrinsic and extrinsic factors triggering autumn leaf senescence, including lagged effects of spring and summer productivity, are poorly understood. Here, we use 434,226 spring leaf-out and autumn leaf senescence observations of temperate trees from Central Europe between 1948 and 2015 to test the effect of seasonal photosynthetic activity on leaf senescence, thereby exploring the extent to which growing-season lengths are internally regulated by constraints on productivity. We found that spring and summer productivity was a critical driver of autumn phenology, with earlier leaf senescence in years with high seasonal photosynthetic activity. Our new process-based model, incorporating information on growing-season photosynthesis, increased the accuracy of existing autumn phenology models by 22–61%. Furthermore, the physiological constraint of growing-season photosynthesis reversed the predictions of autumn phenology over the rest of the century. While current phenology models predict that leaf senescence will occur 7–19 days later by the end of the 21st century, we estimate that leaf senescence will, in fact, advance by 3–6 days. Our results reveal important constraints on future growing-season lengths and the carbon uptake potential of temperate trees and enhance our capacity to forecast long-term changes in ecosystem functioning, which is critical to improve our understanding of Earth System dynamics in response to climate change.
How to cite: Zohner, C.: The effect of growing-season productivity on autumn phenology in temperate trees, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-9937, https://doi.org/10.5194/egusphere-egu21-9937, 2021.
The timing of a tree's leaf emergence represents a trade-off between maximising competition for resources (e.g. light, nutrients) and avoiding freezing damage. Global warming has significantly advanced the date of the last frost events in temperate zones, but in parallel has also shifted the onset of vegetation in spring over the last decades. Thus, the risk of frost damage to plants has not necessarily decreased, depending on geographical location and species. In this study we aim to assess the overall impact of frost damage for saplings vitality. We used saplings of 4 temperate, deciduous tree species (Prunus avium, Carpinus betulus, Quercus petrea and Fagus sylvatica) and artificially altered the leaf-out date by applying a warming or cooling treatment before the natural leaf-out to reflect the whole range of possible leaf-out dates. Once leaves emerged, we simulated a natural frost event, damaging all or half of the new leaves. We then analyzed how fast the different species recovered depending on leaf-out timing in terms of recovery time (time until second flush), growth (biomass and heigth) and non-structural carbohydrate reserves (NSC) in relation to non-frozen control plants. By quantifying the penalty of frost damages in late spring this experiment aims to specify the risk of a species’ strategy to time spring phenology.
How to cite: Baumgarten, F., Gessler, A., and Vitasse, Y.: The penalty of spring frost damages from earliest to latest possible leaf-out timings in temperate forest trees, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-16132, https://doi.org/10.5194/egusphere-egu21-16132, 2021.
Plant phenology is mainly driven by temperature in extratropical ecosystems. Contrasting responses of foliar phenology to climatic warming, however, have been reported in recent decades, raising important questions about the role of other environmental constraints, especially light. A striking and common aspect to past phenological studies is that all analyses have been solely based on air temperature. In fact, temperatures differ substantially between plant tissues and the air, because plants absorb and radiate energy. Using a simple model of bud energy balance, we explore how using bud instead of air temperature could change our interpretation of the phenological response to warming and explain several observed responses of phenology to temperature and light. Not accounting for the real temperature of plant tissues represents a real gap in phenology studies. Field observations of plant tissues temperature as well as experiments are needed for accurately assessing the response of vegetation to climate change.
How to cite: Peaucelle, M., Peñuelas, J., and Verbeeck, H.: Phenology studies need to account for tissue temperature, not air, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-5, https://doi.org/10.5194/egusphere-egu21-5, 2020.
Budbreak is one of the most observed and studied phenological phases in perennial plants. Historically, two effects of temperature are used to model budbreak: the accumulation of heat units (forcing); and the accumulation of time spent at low temperatures (chilling). These two effects have a well-established negative correlation: the more chilling, the less forcing is required to reach budbreak. However, prediction of budbreak remains a challenge, as even artificial warming experiments do not match changes in observed budbreak timing during the past few decades of climate warming. The cold hardiness of buds is, however, largely ignored in estimations of timing to budbreak. Cold hardiness level fluctuates throughout the winter as temperatures change, constantly altering the initiation point of deacclimation. During budbreak assays, cold hardiness loss is extremely slow (low deacclimation rate) at low chill accumulation, and increases to a maximum at high chill accumulation. By standardizing deacclimation rates for each species based on the maximum observed, a deacclimation potential describes dormancy fulfillment. Our studies show that deacclimation rates vary at different temperatures demonstrating the effect of forcing is non-linear. We show that the concept of variable chilling requirements for satisfying dormancy (high chill vs. low chill) is largely erroneous and instead these phenotypes reflect previously unmeasured differences between species or genotypes regarding the interaction between cold hardiness state and deacclimation potential. Our studies show that forcing responses (maximum rates of deacclimation) are normally distributed within a species, and are a heritable trait. Three effects of temperature are thus necessary to describe contemporary phenology patterns as well as predict future impacts of climate change: the accumulation of chill, the forcing temperature response, and the cold hardiness of buds.
How to cite: Kovaleski, A. and Londo, J.: If you work with phenology, you work with bud cold hardiness dynamics, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-6761, https://doi.org/10.5194/egusphere-egu21-6761, 2021.
During winter dormancy in deciduous species, water stops flowing in the xylem and buds become isolated from the stem xylem conduits by a physical barrier made of callose deposits. During bud break, the plant builds new vascular connections between the growing buds and the xylem to support sap flow and transpiration in the developing leaves. However, little is known about the exact timing when these new vascular connections are made, or about the origin of the water supporting bud swelling prior to bud break. This information is particularly limited in forest tree species. We aimed to clarify the origin of the water entering the buds at different developmental stages in temperate forest tree species using water stable isotope tracing techniques to track water movement between soil, stem and buds. More specifically, we developed a method to collect sap water separately from water in other stem tissues (Barbeta et al. 2020). At different leaf phenological stages during the 2018 growing season, we collected soil, stem, bud and leaf samples from 5 adult trees and 3 species (Fagus sylvatica, Quercus robur, Pinus pinaster) growing in a riparian forest in Southwest France. We estimated the relative water content in each sample by extracting bulk water by cryogenic vacuum distillation, and also extracted sap water from stem samples using our new method. All water samples were then analysed for their stable isotope composition (δ18O, δ2H). These results, complemented by some additional labelling experiments, provide key information about the timing of hydraulic reconnection between the buds and the xylem and about the source of water supporting bud swelling and bud break, demonstrating the usefulness of water stable isotope measurements to understand water transport pathways during bud development and canopy leaf out.
Barbeta, A., Burlett, R., Martín-Gómez, P., Fréjaville, B., Devert, N., Wingate, L., Domec, J.-C., et al. (2020). Evidence for distinct isotopic composition of sap and tissue water in tree stems: Consequences for plant water source identification. BioRxiv, 2020.06.18.160002.
How to cite: Martín Gómez, P., Ogée, J., Burlett, R., Barbeta, A., Devert, N., Dubois, S., Frejaville, B., and Wingate, L.: Where do the buds get their water from during budburst? New perspectives from temperate forest species using water stable isotopes, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-8272, https://doi.org/10.5194/egusphere-egu21-8272, 2021.
Leaf-out of deciduous trees is regulated by a set of environmental factors such as cool temperatures during winter-dormancy (chilling), warm spring temperatures (forcing), and daylength (photoperiod), with complex interactions between these factors. Teasing apart these different factors in situ is challenging as no visible changes occurs during the dormancy phase. Manipulating these factors in climate chamber experiments may overcome this issue but may not reflect how they truly interact in natural conditions. Previous researches suggested that bud meristems are disconnected from the xylem flow during endodormancy and that the connection become progressively restored once exposed to a certain duration of chilling. Here we developed a new method using isotopically labelled water (D2O) to quantify the amount of water that can reach buds during the whole dormancy till budburst for 5 different species (Acer pseudoplatanus, Carpinus betulus, Fagus sylvatica, Quercus petraea, Tilia cordata).
In detail, we harvested twig cuttings from leaf fall to budburst (~every two weeks, 12 times) of these species from two different sites (about 5°C of difference) and placed them into labelled water during 24 h at constant light and 20°C. Buds were then cut and water content extracted to quantify δD. Thus, tracing back the water flow into the buds by the amount of D2O taken up. In parallel a subset of twigs was left in the room at 20°C to assess the time to budburst as a proxy for dormancy depth. Analyses of the data are ongoing and preliminary results show progressive increase of water uptake after induction of winter dormancy until budburst as chilling duration increased. Further, we also found distinct differences between species whereas Carpinus betulus showed the highest and Tilia cordata the lowest label uptake during winter dormancy. Furthermore, individuals growing at higher elevation took up less label indicating a stronger dormancy at lower winter temperatures. In summary, we think that our method seems a valuable tool to track quantitative changes in dormancy depth of temperate species especially, in combination with investigations on the molecular level such as sugars or hormones during winter-dormancy.
How to cite: Walde, M. G., Saurer, M., and Vitasse, Y.: Isotopically labelled water - A valuable tracer to track initiation and progress of bud dormancy in temperate trees, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-15904, https://doi.org/10.5194/egusphere-egu21-15904, 2021.
Winter leaf reddening is a phenomenon that evergreen species’ leaf color changes into red resulting from the accumulation of red pigments before or during winter, which persists for several months before dissipating with springtime warming. Among the many hypotheses about the winter leaf reddening, photoprotection is currently the favored hypothesis. Several studies focused on leaf reddening in angiosperms species. Yet, little researches concerned about leaf reddening in gymnosperms species. In gymnosperms, a kind of xanthophyll pigment rhodoxanthin was reported to play an important role. However, the xanthophyll cycle is the main protection mechanism of plants to deal with excessive light energy.
To track the winter leaf reddening phenomenon, we utilized the carotenoid-based vegetation index, the photochemical reflectance index (PRI), which is sensitive to changes in carotenoid pigments (e. g. xanthophyll pigments) in live foliage, as a tool to reflect the invisible phenology of photosynthesis by assessing carotenoid pigment dynamics. We used the CO2 flux data and the micrometeorological data collected from the temperate Japanese cypress forest from 2014 to 2019. We also made use of the digital camera to monitor the canopy phenology changes from 2016 to 2019. The digital camera took photos in 3 hours intervals with 3 different ROI (region of interest), the RGB channels of image data were extracted to calculate the RGB chromatic coordinates and the Red-Green vegetation index (RGVI).
Our findings demonstrated that air temperature reached the lowest point had a one-month lag in the time than that of PAR. The imbalance between light energy absorption and light energy utilization might activate the photoprotection mechanism. The change in light use efficiency (LUE) might confirm this conjecture. LUE reached its peak at the end of December and then dropped sharply. It suggested the photoprotection mechanism was activated. The RGVI fluctuation showed the seasonal changes with that of PRI almost in contrast. PRI was highly correlated with RGVI (R=-0.806928034317071 in Pearson’s correlation test). It suggested that the winter leaf reddening phenomenon caused the decline of PRI. Further, the PRI and RGVI both were highly correlated with air temperature and PAR. Based on current observations, there are still many unclear mechanisms. In the future, we will try to better explain the mechanism of winter reddening with a new set of experiments.
Keyword: winter leaf reddening, Japanese cypress, photochemical reflectance index (PRI), Red-Green vegetation index (RGVI), phenological analysis, digital camera
How to cite: Chen, S., Kosugi, Y., Jiao, L., Nakaji, T., Noda, H., Hikosaka, K., and Nasahara, K.: Winter leaf reddening phenomenon: the long-term track of PRI and phenological changes in a temperate Japanese cypress forest at Kiryu Japan., EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-6849, https://doi.org/10.5194/egusphere-egu21-6849, 2021.
Phenological models applied to grapevines are valuable tools to assist in the decision of cultural practices related to winegrowers and winemakers. The two-parameter sigmoid phenological model was used to estimate the three main phenological stages of the grapevine development, i.e., budburst, flowering, and veraison. This model was calibrated and validated with phenology data for 51 grapevine varieties distributed in four wine regions in Portugal (Lisboa, Douro, Dão, and Vinhos Verdes). Meteorological data for the selected sites were also used. Hence, 153 model calibrations (51 varieties × 3 phenological stages) and corresponding parameter estimations were carried out based on an unprecedented comprehensive and systematized dataset of phenology in Portugal. For each phenological stage, the centroid of the estimated parameters was subsequently used, and three generalized sigmoid models were constructed (budburst: d =−0.6, e = 8.6; flowering: d = −0.6, e = 13.7; veraison: d = −0.5, e = 13.2). Centroid parameters show high performance for approximately 90% of the varieties and can thereby be used instead of variety-specific parameters. Overall, the RMSE (root-mean-squared-error) is < 7 days, while the EF (efficiency coefficient) is > 0.5. Additionally, according to other studies, the predictive capacity of the models for budburst remains lower than for flowering or veraison. Furthermore, the F-forcing parameter (thermal accumulation) was evaluated for the Lisboa wine region, where the sample size is larger, and for the varieties with model efficiency equal to or greater than 0.5. A ranking and categorization of the varieties in early, intermediate, and late varieties was subsequently undertaken on the basis of F values. In this way, these results of the present study will be incorporated on a web platform, where the sigmoid model must convey valuable information regarding the development/evolution of the vineyard with short-term predictions.
Keywords: grapevine; phenology modeling; sigmoid model; wine regions; short-term predictions; Portugal
How to cite: Reis, S., Fraga, H., Carlos, C., Silvestre, J., Eiras-Dias, J., Rodrigues, P., and A. Santos, J.: Grapevine Phenology in Four Portuguese Wine Regions: Modeling and Predictions, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-671, https://doi.org/10.5194/egusphere-egu21-671, 2021.
According to the Köppen climate classification, almost the entire area of Latvia belongs to the same climate type, Dfb, which is characterized by humid continental climates with warm (sometimes hot) summers and cold winters. In the last decades whether conditions on the western coast of Latvia more characterized by temperate maritime climates. In this area there has been a transition (and still ongoing) to the climate type Cfb.
Temporal and spatial changes of temperature and precipitation regime have been examined in whole territory to identify the breaking point of climate type shifts. We used two type of climatological data sets: gridded daily temperature from the E-OBS data set version 21.0e (Cornes et al., 2018) and direct observations from meteorological stations (data source: Latvian Environment, Geology and Meteorology Centre). The temperature and precipitation regime have changed significantly in the last century - seasonal and regional differences can be observed in the territory of Latvia.
We have digitized and analysed more than 47 thousand phenological records, fixed by volunteers in period 1970-2018. Study has shown that significant seasonal changes have taken place across the Latvian landscape due to climate change (Kalvāne and Kalvāns, 2021). The largest changes have been recorded for the unfolding (BBCH11) and flowering (BBCH61) phase of plants – almost 90% of the data included in the database demonstrate a negative trend. The winter of 1988/1989 may be considered as breaking point, it has been common that many phases have begun sooner (particularly spring phases), while abiotic autumn phases have been characterized by late years.
Study gives an overview aboutclimate change (also climate type shift) impacts on ecosystems in Latvia, particularly to forest and semi-natural grasslands and temporal and spatial changes of vegetation structure and distribution areas.
This study was carried out within the framework of the Impact of Climate Change on Phytophenological Phases and Related Risks in the Baltic Region (No. 126.96.36.199/VIAA/2/18/265) ERDF project and the Climate change and sustainable use of natural resources institutional research grant of the University of Latvia (No. AAP2016/B041//ZD2016/AZ03).
Cornes, R. C., van der Schrier, G., van den Besselaar, E. J. M. and Jones, P. D.: An Ensemble Version of the E-OBS Temperature and Precipitation Data Sets, J. Geophys. Res. Atmos., 123(17), 9391–9409, doi:10.1029/2017JD028200, 2018.
Kalvāne, G. and Kalvāns, A.(2021): Phenological trends of multi-taxonomic groups in Latvia, 1970-2018, Int. J. Biometeorol., doi:https://doi.org/10.1007/s00484-020-02068-8, 2021.
How to cite: Kalvāne, G., Kalvāns, A., Briede, A., Krampis, I., Kaupe, D., and Rūsiņa, S.: The shifting of climate types: manifestation to phenology and ecosystems structure, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-5689, https://doi.org/10.5194/egusphere-egu21-5689, 2021.
Traditional phenological models use the concepts of chilling and thermal forcing (temperature sum or degree-days) to predict buds break. Even if new model formulations get more sophisticated with time, the bases of phenological model still rely on the effect of the time of chilling and forcing temperature in interaction, or not, with photoperiod. Because of the increasing impact of climate or other related biotic or abiotic stressors, a model with more biological support is urgently needed in order to accurately predict bud break. We have developed and calibrated a new mechanistic model that is based on the physiological processes taking place before and during budbreak in several conifers species. This model describes the phenology and growth dynamics of a conifer branch as representative of the whole tree. As a general assumption, we assume that phenology will be driven by the carbon status, which is closely related to the annual cycle of dormancy – activity state through the year and to the environmental variables. The carbon balance of a branch was thus modelled i) from autumn to winter–when aboveground parts exhibit cold acclimation and dormancy– and ii) from winter to spring and summer –when deacclimation and growth resumption occurs. After being calibrated in a field experiment, the model was tested across a large area in Québec (Canada), based on observed phenological data. For the 20 field sites in Quebec, the model proved to be accurate in predicting the date of budbreak with an average error of ±3.8 days (R2=0.72). This model also allowed us to better understand the effects of winter and spring temperature on bud burst, offering new simulation perspectives under global warming and insect defoliation.
How to cite: Deslauriers, A., Carteni, F., Balducci, L., Dupont, A., and Mazzoleni, S.: Development of a new phenological model based on the carbon balance of tree in boreal conifers, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-10272, https://doi.org/10.5194/egusphere-egu21-10272, 2021.
In perennial grasses, the reproductive development encompasses several phenological events, such as apex induction, floral transition, heading and flowering, that deeply affect biomass production, forage quality and plant perenniality. Despite the importance of perennial grasses in agricultural systems and natural ecosystems, we still lack accurate models predicting the reproductive development and its consequences on plant growth and grassland management. Most of available models implements a fixed scheduling of the reproductive development expressed either in thermal time or in calendar time. The progressive completion of floral induction and the effects of environmental factors are generally poorly described. In addition, the vegetative and reproductive developments are represented as independent and successive phases. In the present work, we introduce the new model LgrassFlo, which simulates the reproductive development of perennial grasses in interaction with plant vegetative development and considering the effects of environmental conditions on floral induction.
LgrassFlo simulates the canopy as the dynamics of a collection of individual plants, each being composed of one or more tillers. The 3D description of leaf growth and tillering is based on a functional-structural plant model of perennial ryegrass (Lgrass). We developed a new model of floral induction describing the progression of the primary and secondary induction of each apex of the plant according to (i) the daily temperature, (ii) photoperiod and (iii) plant architecture. This model was coupled to Lgrass, the model ensemble being called LgrassFlo. During apex induction, LgrassFlo accounts for an increase in the rates of leaf primordia initiation and leaf elongation. After floral transition, we assume that the apex only initiates spikelet primordia and that internodes start to elongate. LgrassFlo simulates the date of floral transition, the final number of leaves and the heading date based on a 3D representation of plant architecture.
A specific experiment was carried out in order to calibrate LgrassFlo on data describing the vegetative and reproductive development of three Lolium perenne cultivars contrasted for their precocity and exposed to four inductive conditions in growth chambers. The first three conditions consisted in a period allowing for primary induction (low temperature – short day) followed by a period allowing for secondary induction (high temperature – long day), the two periods being spaced by a non-inductive period (high temperature - short day) of 0, 3 or 6 weeks. In the fourth condition, plants were not exposed to conditions allowing for the primary induction. A set of vegetative and reproductive parameters were estimated for each individual plant of the experiment. The parameter values were independent of the experimental treatment but showed a large genetic diversity both between and within varieties. Using this calibration, LgrassFlo satisfactorily predicted the observed diversity in final leaf number and heading date.
The present model is a step forward towards a better prediction of perennial grass phenology in actual and future climatic conditions. In this respect, the model is being currently used to simulate the observed genetic diversity in the heading date of several Lolium perenne cultivars grown in contrasted temperate climates over the last 15 years.
How to cite: Rouet, S., Durand, J.-L., Combes, D., Escobar-Gutierrez, A., and Barillot, R.: Calibration of LgrassFlo, a new model of perennial grass phenology in response to temperature and photoperiod., EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-12430, https://doi.org/10.5194/egusphere-egu21-12430, 2021.
Understanding the impact of the climate crisis on our planet, and hence, on human and natural life, is a pressing but challenging scientific endeavor. Phenological studies help to build such an understanding by analyzing changes in the timing of biological events. Such changes are, in turn, linked to changes in the likelihood of experiencing false springs. Here we examine the risk and uncertainty of false springs over Europe using the extended spring indices models and the E-OBS weather dataset (1950-onwards). False spring uncertainty is evaluated using the full ensemble of 100 daily weather realizations that accompany the E-OBS dataset. Smart computing is used to handle this relatively large amount of gridded data. Our results show that changes in false spring risk are heterogeneous in space, with increasing risks mostly found in the mid-latitudes and decreasing risks mostly found at high and low latitudes. From 1950 until 1979, there was a substantial increase in false spring risk, whereas the change in false spring risk from 1980 onwards is negligible. Our results also show that false spring uncertainty is highest in western Europe. These results highlight the importance of considering temperature uncertainty in phenological modeling, especially when examining the risk of false springs.
How to cite: Zurita-Milla, R., Vermeltfoort, R., Girgin, S., and Izquierdo-Verdiguier, E.: False Spring over Europe: risk and uncertainties, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-10298, https://doi.org/10.5194/egusphere-egu21-10298, 2021.
The imprint of recent climate change on plant phenology has been the subject of several investigations during the last few decades. Results of such studies have repeatedly documented the advances of key phenological stages in spring. More recently, they have also shown that global warming has induced changes in temperature sensitivity and led to more uniform phenology across elevations. While awareness of trends in phenology undoubtedly contributes to inform ecosystem management, the provision of ecosystem services also necessitates knowledge and understanding of how spring phenology varies from year to year. For instance, in view of growing exposure of grassland ecosystems to summer drought, in Alpine countries forage production increasingly relies on exploiting at best spring growth, which in turn requires an accurate timing of field operations, depending on the progress of herbage development.
Employing long-term phenological observations on forest trees and grassland plants and weather records from Switzerland, in this contribution I examine year-to-year variations in spring phenology in light of anomalies in the seasonal mean temperature for the months of February to April, and reflect on how the latter can be related to number of dry days and associated temperature anomalies. Based on these findings and results from other studies, I discuss possible implications of future climate change for the variability of spring phenology.
How to cite: Calanca, P.: Dry days, associated temperature anomalies and inter-annual variations in spring phenology, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-12512, https://doi.org/10.5194/egusphere-egu21-12512, 2021.
“Phenology – the timing of seasonal activities of animals and plants – is perhaps the simplest process in which to track changes in the ecology of species in response to climate change” (IPCC 2007).
PEP725, the Pan-European Phenological Database, is a European research infrastructure to promote and facilitate phenological research. Its main objective is to build up and maintain a European-wide phenological database with an open, unrestricted data access for science, research and education. So far, 20 European meteorological services and 6 partners from different phenological network operators have joined PEP725.
The PEP725 phenological data base (www.pep725.eu) now offers more than 12 million phenological observations, all of them classified according to the so called BBCH scale. The first datasets in PEP725 date back to 1868; however, there are only a few observations available until 1950. Having accepted the PEP725 data policy and finished the registration, the data download is quick and easy and can be done according to various criteria, e.g., by a specific plant or all data from one country. The integration of new data sets for future partners is also easy to perform due to the flexible structure of the PEP725 database as well as the classification of the observed plants via the so-called gss format (genus, species and subspecies).
PEP725 is funded by EUMETNET, the network of European meteorological services, ZAMG, who is the acting host for PEP, and the Austrian ministry of education, science and research.
The phenological data set has been growing by about 100000 observations per year. Also the number of user registrations has continually been increasing, amounting to 305 new users and more than 28000 downloads in 2020. The greatest number of users are found in China, followed by Germany and the US. To date we could count 78 reviewed publications based on the PEP725 data set with 18 in 2020 and a total of 9 published in Nature and one in Science.
The data base statistics demonstrate the great demand and potential of the PEP725 phenological data set, which urgently needs development including a facilitated access, gridded versions and near real time products to attract a greater range of users.
How to cite: Ressl, H., Scheifinger, H., Hübner, T., Paul, A., and Ungersböck, M.: PEP725 a European phenological database, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-2179, https://doi.org/10.5194/egusphere-egu21-2179, 2021.
Global warming increases the need for local climatic studies in wine-producing areas. Winegrowers have to develop strategies to adapt their activities to new climatic conditions and to their various effects on vine culture. Among them, distribution and population dynamics of pest species are likely to change. New species could reach the temperate regions, and some native species could create more damages than previously in the vineyards. In Western Europe, the distribution of the American grapevine leafhopper Scaphoideus titanus has been observed to shift northwards during the last decades (Boudon and Maixner 2007). Plurivoltin species such as the European grapevine moth Lobesia botrana could produce more generations per year (Gutierrez et al. 2018), creating potentially more damages on grapes. To help winegrowers, it is crucial to lead research at local scale, taking into account microclimatic specificities of the vineyards (Mozell and Thach 2014).
In this study, we examine temperature trends during the growing season in the region of Neuchatel and their potential impacts on major vine pest species. We focus on the American grapevine leafhopper and on the European grapevine moth. The American grapevine leafhopper is already established in the Lake Geneva area and could soon reach the Neuchatel area, while the European grapevine moth is already present in the Neuchatel vineyard. We use temperature data over the last 40 years (1980-2019) and two climatic scenarios to assess present suitability for pest development and the perspectives for the next decades.
Boudon, E. & M. Maixner. 2007. Potential effects of climate change on distribution and activity of insect vectors of grapevine pathogens. In International and multi-disciplinary" Global warming, which potential impacts on the vineyards?".
Gutierrez, A. P., L. Ponti, G. Gilioli & J. Baumgärtner (2018) Climate warming effects on grape and grapevine moth (Lobesia botrana) in the Palearctic region. Agricultural and Forest Entomology, 20, 255-271.
Mozell, M. R. & L. Thach (2014) The impact of climate change on the global wine industry: Challenges & solutions. Wine Economics and Policy, 3, 81-89.
How to cite: Schneider, L., Comte, V., Sneiders, B., and Rebetez, M.: Climate change in the vineyard: perspectives for pest species in the region of Neuchatel, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-16335, https://doi.org/10.5194/egusphere-egu21-16335, 2021.
Emissions of biogenic aerosols such as allergenic pollen affect the public health badly. In combination with air pollution it puts additional distress on people already suffering from cardiovascular and respiratory diseases. In some European countries the prevalence of people with pollinosis is up to 40%. In Belgium, ~10% is sensitive for birch pollen. Patients suffering from pollinosis in Belgium lack access to detailed real-time spatial information and warnings on forthcoming pollen exposures. This is because the only pollen info is coming from five aerobiological stations which monitor off-line daily concentrations of airborne pollen from birches. Only two stations have almost four decades of observations, Brussels from 1982 on and De Haan from 1984 on. Chemistry Transport Models (CTM) can both quantify as well as forecast the spatial and temporal distribution of airborne birch pollen concentrations if the distributions of birch pollen emission sources over time are available.
Here we show the results of the modelled spatio-temporal distributions of almost four decades of birch pollen levels over Belgium using the CTM SILAM (http://silam.fmi.fi). This CTM is driven with the ERA5 meteorological reanalysis from ECMWF, and reconstructed birch tree fraction maps. A recent in-house birch map of Belgium derived from forest inventory data is combined with long-term series of the AVHRR-GIMMS3g NDVI to produce birch fraction maps for each year.
For the first time in Belgium, we present time series of modelled birch pollen levels by SILAM compared with daily observations from the aerobiological surveillance network for the period 1982-2019. Preliminary modelling results for Brussels show an overall R² value of 0.40 computed from modelled and observed daily birch pollen levels. The R² values for the individual birch pollen seasons may range from 0.10 to 0.82 with a median value of 0.53. For De Haan the R² values tend to be lower with the median seasonal value of 0.30. Temporal trends computed on the first results of the modelled daily values based on the Theil Sen slope and the Area Under the Curve (AUC) show a substantial increase of birch pollen levels for most parts in Belgium. This agrees well with literature reports.
How to cite: Verstraeten, W., Bruffaerts, N., Hoebeke, L., Kouznetsov, R., Sofiev, M., and Delcloo, A.: Modelling the spatial distribution of four decades of airborne birch pollen levels in Belgium using remotely sensed birch fraction maps, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-1103, https://doi.org/10.5194/egusphere-egu21-1103, 2021.
Aeroallergens contribute a major climate change impact on human health since warming favours the production and advances the release of plant pollen. This goes in line with a widely observed advance of flowering in response to increasing temperatures. However, documented plant phenological changes vary with species traits, seasons, and sites. Nevertheless, the start and end of flowering dates are known to build a solid baseline for assessing the spatial and temporal patterns in pollen calendars. A closer look at the match/mismatch of flowering and start of pollen season dates reveals considerable differences which may be also indirectly linked to climate change. In this talk, we will present three perspectives related to (1) grassland land use, cutting regimes and agri-environment measures (AEM), (2) post-season pollen transport of an alpine Alnus species, as well as (3) a first climatology of pre-season long-range pollen transport to Bavaria. These selected examples underline the prominent role of land use/land cover (LULC) and pollen transport besides direct temperature mediated climate change effects on flowering for regional pollen calendars.
How to cite: Menzel, A. and Yuan, Y.: The role of land cover, land use, and atmospheric transport for the mismatch of flowering and atmospheric pollen seasonality, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-2204, https://doi.org/10.5194/egusphere-egu21-2204, 2021.
The prevalence of pollen allergy is increasing worldwide, as is the proportion of people living in cities. Thus, there is an increasing importance to investigate pollen distribution across city districts. We conducted two sampling campaigns to investigate the spatial and temporal variation of airborne pollen in the Sydney metropolitan area and the vertical variation within a forest in north-western Sydney.
Spatial assessment of pollen deposition was made for eight weeks in the exceptionally dry summer in 2019/2020 using gravimetric samplers. These samplers were set up at ten locations characterised by different degrees of urbanisation and distance to the sea. We focussed on the most abundant pollen types and investigated statistical relationships with land use and meteorology. In addition, we compared our results with pollen data of previous years sampled at a pollen monitoring station located in north-western Sydney in a semi-rural environment. We measured vertical pollen concentrations in a native forest, which mostly consists of Eucalypt trees (family Myrtaceae) in north-western Sydney. A scaffolding was equipped with five portable volumetric pollen samplers installed at different heights (1, 4, 10, 16, 20.5 m above ground level (agl)). We measured pollen concentration every second hour between 9 am and 4 pm on a total of four days in January 2020. We compared concentrations between days, heights, and times of the day.
The most abundant pollen type registered within our sampling campaigns belonged to the family Myrtaceae. Grass pollen (Poaceae) was also detected, but in much smaller quantities which can be attributed to the drought and temporal setting of the campaign, which started in the post-peak period of a comparably weak pollen season associated with a smaller number of days with medical relevance (> 50 pollen grains/m3). Our data showed spatial variations between the ten locations, but no relationship with land use (grass and tree cover) and meteorology could be found. This suggests the influence of other factors such as long-range pollen transport or resuspension of pollen. In the forest, Myrtaceae concentrations varied between days, sampling height and time of the day: the highest concentration was recorded on the second day of measurement between 9 and 10 am at 10 m agl. Peak values were generally reached between 1 and 2 pm. Considering sampling height, concentrations were on average highest at 4 m agl. The location of pollen sources as well as meteorological conditions such as turbulence and variation in wind speed may be key determinants of small-scale differences of pollen concentrations.
The drought preceding this study did not only influence the length but also the strength of the pollen season. Data on vertical variations could support investigations related to turbulence, which is also responsible for resuspension processes.
How to cite: Jetschni, J., Al Kouba, J., Beggs, P. J., and Jochner-Oette, S.: Investigations on spatial and vertical variations of airborne pollen in Sydney, Australia, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-9184, https://doi.org/10.5194/egusphere-egu21-9184, 2021.
Rising CO2 levels and climate change may be resulting in some shift in the geographical range of certain plant species, as well as in increased rate of photosynthesis. Many plants respond accordingly with increased growth and reproduction and possibly greater pollen yields, that could affect allergic diseases among other things.
The aim of this study is the evolution of aerobiological measurements in France for 25-30 years. This allows to follow the main phenological parameters in connection with the pollination and the ensuing allergy risk.
Material and method:
The RNSA (French Aerobiology Network) has pollen background-traps located in more than 60 towns throughout France. These traps are volumetric Hirst models making it possible to obtain impacted strips for microscopic analysis by trained operators. The main taxa studied here are birch, grasses and ragweed for a long period of more than 25 years over some cities of France.
Concerning birch but also other catkins or buds’ trees pollinating in late winter or spring, it can be seen an overall advance of the pollen season start date until 2004 and then a progressive delay, the current date being nearly the same as it was 20 years ago, and an increasing trend in the quantities of pollen emitted.
For grasses and ragweed, we only found a few minor changes in the start date but a longer duration of the pollen season.
As regards the trees, the start date of the new production of catkins or buds is never the 1st of January but depends on the species. For example, it is early July for birch. For breaking dormancy, flowering, and pollinating, the trees and other perennial species need a period of accumulation of cold degrees (Chilling) and later an accumulation of warm degrees (Forcing). With climate change these periods may be shorter or longer depending of the autumn and winter temperature. Therefore, a change in the annual temperature may have a direct effect on the vegetal physiology and hence on pollen release. It may also explain why the quantities of pollen produced are increasing.
The Poaceae reserve, from one place to another and without any spatial structuring, very contrasted patterns which make it impossible to identify a general tendency. This is probably due to the great diversity of taxa grouped under the generic term Poaceae, which are clearly not equally sensitive to climate change.
Trees with allergenic pollen blowing late winter or early spring pollinate since 2004 later and produce amounts of pollen constantly increasing. Grasses and ragweed have longer periods of pollination with either slightly higher or most often lower pollen production.
How to cite: Monnier, S., Thibaudon, M., Besancenot, J.-P., Sindt, C., and Oliver, G.: Phenological phases of pollination related to climate change, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-15034, https://doi.org/10.5194/egusphere-egu21-15034, 2021.
The use of satellite sensors, near-surface cameras and other remote methods of monitoring vegetation phenology at landscape and higher scales has become increasingly common. These technologies provide a means to determine the timing of phenophases and growing season length at different spatial resolutions; coverage that is not attainable by human observers. However, in situ ground observations are necessary to validate remotely derived phenometrics. Despite increased knowledge and expertise there still remains the persistent challenge of reconciling ground observations at the individual plant level with remotely sensed (RS) phenometrics at landscape or larger scales. Here, we compared the timing of in situ phenophase estimates (spring and autumn) with a range of corresponding remote sensing (MODIS, VIIRS, PhenoCam) phenometrics across five terrestrial sites in the USA’s National Ecological Observatory Network (NEON). The sites represent a range of ecosystem types including, deciduous forest (Harvard Forest, MA), dry scrubland (Onaqui, UT), evergreen forest (Abby Road, WA) and seasonal wetlands (Disney Wilderness Preserve and Ordway-Swisher Biological Station, FL) focusing on a three year period from 2017-2019. Our main objective was compare a range of co-located RS pheometrics with in situ observations to explore potential reasons for the observed discrepancies and to determine which technologies were more aligned with ground observations. Statistically significant relationships were strongest (p<0.001) for spring phenophases compared to autumn. In general, satellite derived phenometrics tended to be earlier (RMSE 21.7 – 28.4 days) than in situ spring phenology whereas PhenoCam derived phenometrics were later (RMSE 24 days). Overall, discrepancies between in situ and RS phenometrics related to scale, species availability and the short duration of the time-series (3 years). However, as the NEON project progresses these challenges are expected to be reduced as more data become available.
How to cite: Donnelly, A., Yu, R., Jones, K., Belitz, M., Li, B., Duffy, K., Zhang, X., Wang, J., Seyednasrollah, B., Gerst, K., Li, D., Kaddoura, Y., Zhu, K., Morisette, J., Ramey, C., and Smith, K.: Comparing in situ phenology and remotely derived phenometrics across ecosystems, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-3087, https://doi.org/10.5194/egusphere-egu21-3087, 2021.
Phenology is an important driver of ecosystem performance. However, studies of phenology in Ireland have been limited by the availability of data at high spatial and temporal resolutions. The new suite of Sentinel-2 sensors, with their enhanced spatial and temporal resolutions might help overcome some of these challenges. Additionally, the presence of red edge bands in the Sentinel-2 sensors provides a unique opportunity to evaluate the performance of different vegetation indices in tracking near surface (phenocam) and ground/laboratory measures of phenology. In this study, we present our initial analyses for the year 2020. Nine common lime trees (Tilia x europaea) on the University College Cork campus (Cork, Ireland) and three undisturbed broadleaf woodland sites from the National Park and Wildlife Services (NPWS) survey were selected. The phenology of these sites was analyzed from satellite derived vegetation indices of NDVI, EVI, GNDVI and NDRE. The available 24 cloud free Sentinel-2 images were pre-processed and interpolated to daily time steps. The start of season (SOS), position of peak (POP) and end of season (EOS) were then extracted from the daily time series using the half amplitude and maximum value method. Similarly, daily data from a phenocam overlooking three of the lime trees were processed to extract the phenological dates. Weekly measurements of leaf chlorophyll or chlorophyll content index (CCI) and maximum photosystem II efficiency (Fv/Fm) by sampling five leaves from each lime tree were made during June to November of 2020. Preliminary results indicate that different vegetation indices vary in their correlation to ground and phenocam observations. The dates of SOS, POP and EOS obtained from Sentinel-2 do not exactly match the ground and phenocam observations, nor are the different indices coincident with each other, with maximum deviations of up to a month and a week for EOS and SOS respectively. The phenological metrics estimated from the EVI time series were in general earlier (i.e. 116, 162 and 270 day of year for SOS, POP and EOS respectively) and those from the NDRE were the last (i.e. 131, 211 and 288 day of year for SOS, POP and EOS respectively). Although local differences were observed in the field, the Sentinel-2 time series data were shown to perform well in tracking the autumn phenology, and in most cases the observed mismatches in phenological data could be ascribed to differences in the scale of observations i.e. pixel vs point comparisons and on spectral basis i.e. sensor vs instrument for measuring CCI. A steeper drop in CCI and Fv/Fm values was also observed in the late autumn period. Such differences in the progression of each time series curve can possibly lead to mismatches in the phenology estimated from vegetation indices and from observations. Other mismatches could also emanate from the fact that field sampling of leaves was done from below the canopy whereas the satellite view of canopy is from the top. Experience from the field revealed differences in the rates of greening and yellowing of the leaves in different regions of the tree canopy.
How to cite: Misra, G., Cawkwell, F., and Wingler, A.: Monitoring deciduous tree phenology estimates with Sentinel-2, phenocam and field measurements in Ireland, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-8038, https://doi.org/10.5194/egusphere-egu21-8038, 2021.
In the semi-arid Peruvian Andes, the agricultural growing season is mostly determined by the timing of the onset and cessation of the wet season, to which annual crop yields are highly sensitive. Recently, local farmers in the Rio Santa valley (Callejón de Huaylas) bordered by the glaciated Coordillera Blanca to the east and the unglaciated Coordillera Negra to the west, reported increasing challenges in the predictability of the onset, more frequent dry spells and extreme precipitation events during the wet season. Previous studies based on time-series of local rain gauges however did not show any significant changes in either timing or intensity of the wet season. Both in-situ and satellite rainfall data for the region lack the necessary spatial resolution to capture the highly variable rainfall distribution typical for complex terrain, and are often of questionable quality and temporal consistency. As in other Andean valleys, there remains considerable uncertainty in the Rio Santa basin regarding hydrological changes over the last decades. These changes are of a great concern for the local society and the lacking knowledge about changes in water availability (i.e. rainfall) and water demand (i.e. land use practices) hinder the assessment of relevant factors for the development of adaption strategies.
The over-archiving goal of this study was to better understand variability and recent changes of plant growth and rainfall seasonality and the interactions between them in the Rio Santa basin. Specifically, we aimed to illustrate how satellite-derived information on vegetation greenness can be exploited to infer a robust and highly resolved picture of recent changes in rainfall and vegetation across the region: As the semi-arid climate causes water availability (i.e. precipitation) to be the key limiting factor for plant growth, patterns of precipitation occurrence and the seasonality of vegetation indices (VIs) are tightly coupled. Therefore, these indices can serve as an integrated proxy of rainfall. By combining a 20 year time series of MODIS Aqua and Terra VIs (from 2000 to today) and datasets of precipitation (both remote-sensing and observations) we explore recent spatial and temporal changes in vegetation and water availability by combining VIs timeseries and derived land surface phenology (LSP) with measures of wet season onset and cessation from rainfall data. Furthermore, we analyse the interaction of El Niño Southern Oscillation (ENSO) and the wet and growing season.
We find spatially variable but significant greening over the majority of the Rio Santa valley domain. This greening is particularly pronounced during the the dry season (Austral winter) and indicates an overall increase of plant available water over time. The start of the growing season (SOS) is temporally highly variable and dominates the variability of growing season length over time. Peak and end of season (POS, EOS) are significantly delayed in the 20 year analysis. By partitioning the results into periods of three stages of ENSO (neutral, Niño, Niña) we find an earlier onset of the rainy and growing season and an overall increased season length in years associated with El Niño.
How to cite: Hänchen, L., Klein, C., Maussion, F., Gurgiser, W., and Wohlfahrt, G.: Vegetation indices as a proxy for spatio-temporal variations in water availability in the semi-arid Rio Santa valley (Callejón de Huaylas, Peru), EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-8330, https://doi.org/10.5194/egusphere-egu21-8330, 2021.
Changing environmental conditions have significantly altered the phenology, spatial distribution, and abundances of species in terrestrial and freshwater ecosystems. Recent work has shown that such changes may alter the strengths of interactions between species and may jumble structural patterns in networks of trophic, mutualistic and/or other interactions that are crucial for biodiversity. ‘Blue’ (aquatic) and ‘green’ (terrestrial) ecosystems are closely interlinked through biogeochemical cycles and species that inhabit both ecosystems. When the effects of abiotic drivers of global environmental change, such as a change in temperature, precipitation, or land use, are different for lakes and their surrounding watersheds, a blue-green phenological mismatch may therefore occur. In particular, because such changes in seasonal patterns may cascade down food webs.
Remote sensing provides spatially and temporally dense information on biochemical properties of the Earth surface, including biomass and primary production indicators for both aquatic and terrestrial ecosystems. Deriving phenology metrics for these indicators is routine practice for terrestrial vegetation, and several case studies demonstrate the feasibility of analogous metrics for lakes and inland seas. In this study, we used remote sensing data to extract phenology metrics (e.g. start of the growing season) for 4264 lakes distributed across a wide range of biomes from daily chlorophyll estimates and vegetation indices spanning a time-period of 15-20 years. We investigate whether changes in the phenology of lake phytoplankton and the surrounding terrestrial vegetation have occurred during this period, and how the phenology in either ecosystem type is synchronized.
Analysis are underway, but preliminary results suggest contrasting results across different biomes as well as substantial differences in the way in which the phenology of primary producers in lakes and on their surrounding watersheds has changed within these biomes.
How to cite: Lever, J., Vitasse, Y., Gilarranz, L. J., D'Odorico, P., and Odermatt, D.: Global remotely sensed phenology of Blue-Green Ecosystems, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-10584, https://doi.org/10.5194/egusphere-egu21-10584, 2021.
The aim of ESA's forthcoming FLuorescence EXplorer (FLEX) is to achieve a global monitoring of the vegetation's chlorophyll fluorescence by means of an imaging spectrometer, FLORIS. For the retrieval of the fluorescence signal measured from space, other vegetation variables need to be retrieved simultaneously, such as (1) Leaf Area Index (LAI), (2) Leaf Chlorophyll content (Cab), and (3) Fractional Vegetation cover (FCover), among others. The undergoing SENTIFLEX ERC project has already demonstrated the feasibility to operationally infer these variables by hybrid retrieval approaches, which combine the generalization capabilities offered by radiative transfer models (RTMs) and computational efficiency of machine learning methods. Reflectance spectra corresponding to a large variety of canopy realizations served as input to train a Gaussian Process Regression (GPR) algorithm for each targeted variable. Following this approach, sets of GPR retrieval models have been trained for Sentinel-2 and -3 reflectance images.
In that direction, we started to explore the potential of Google Earth Engine (GEE) to facilitate regional to global mapping. GEE is a platform with multi-petabyte satellite imagery catalog and geospatial datasets with planetary-scale analysis capabilities, which is freely available for scientific purposes. Among the different EO archives, it is possible to access the whole collection of Sentinel-2 ground reflectance data. In this work, we present the results of an efficient implementation of the GPR-based vegetation models developed for Sentinel-2 in the framework of SENSAGRI H2020 project in GEE. By taking advantage of GEE cloud-computing power, we are able to avoid the typical bottleneck of downloading and process large amounts of data locally and generate results of GPR-based retrieval models developed for Sentinel-2 in a fast and efficient way, covering large areas in matter of seconds. As a first step in that direction we present here an open web-based GEE application able to generate LAI Green and LAI Brown maps from Sentinel-2- imagery at 20m in a tile-wise manner all over the world, and time series of selected pixels during user-defined time interval.
To illustrate this functionalities and have better understanding of the phenology, we targeted a region in Castilla y León (Spain) from where we will present results for 2018 classified per crop type. This land cover classification was generated by the ITACYL (Instituto Tecnológico Agrario de Castilla y León) during SENSAGRI.
Future development will tackle the possibility to extend our analysis capability to additional variables, such as FCover and Cab, maintaining the computational efficiency as the main driver to ensure that the GEE application continues to be an agile and easy tool for spatiotemporal Earth observation studies.
How to cite: Salinero Delgado, M., Pipia, L., Amin, E., Belda, S., and Verrelst, J.: Mapping vegetation variables in Google Earth Engine using Gaussian Process Regression models. , EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-12359, https://doi.org/10.5194/egusphere-egu21-12359, 2021.
Monitoring of crop phenology significantly assists agricultural managing practices and plays an important role in crop yield predictions. Multi-temporal satellite-based observations allow analyzing vegetation seasonal dynamics over large areas by using vegetation indices or deriving biophysical variables. The Northern Nile Delta represents about half of all agricultural lands of Egypt. In this region, intensifying farming systems are predominant, which translates into a pressure on water supply demand. Moreover, double cropping rotations schemes are increasing, requiring a high temporal and spatial resolution monitoring for capturing successive crop growth cycles. This study presents a framework for crop phenological characterization based on high spatial and temporal resolution time series of green Leaf Area Index (LAI). Particularly, NASA's Harmonized Landsat 8 and Sentinel-2 (HLS) surface reflectance dataset was used. The HLS dataset provides seamless products from both satellites, enabling global land observations every 2-3 days at 30m. A green LAI retrieval model was originally trained using ground-based LAI measurements with Gaussian processes technique and validated for Sentinel-2 (R2: 0.7, RMSE= 0.67m2/m2) (Amin et al., 2020). Given the compatible spectral bands configuration of both sensors, a new model for Landsat 8 was adapted from the original one. Both models were implemented in an HLS image based automated retrieval chain obtaining therefore two different LAI time series, which were spatially averaged per crop parcel according to the ground data at disposal. The subsequent analysis was performed based on the time series phenological pre-processing and modelling implemented in the in-house developed scientific time series toolbox DATimeS (Belda et al., 2020). The proposed framework permitted to determine the crop patterns for four consecutive years (2016-2019), identifying one or two seasons per year, for single (e.g. grape, citrus) or double-cropping (e.g. maize-onion, maize-wheat, rice-clover), respectively. Alongside, each detected crop was characterized by retrieving a selected set of phenological parameters, which were contrasted with respect to the established crop type calendar (planting and harvesting dates) and for each crop type, the annual mean value was computed and the intra annual variability within the four years was assessed.
Amin, E., Verrelst, J., Rivera-Caicedo, J. P., Pipia, L., Ruiz-Verdú, A., & Moreno, J. (2020). Prototyping Sentinel-2 green LAI and brown LAI products for cropland monitoring. Remote Sensing of Environment, 112168.
Belda, S., Pipia, L., Morcillo-Pallarés, P., Rivera-Caicedo, J. P., Amin, E., De Grave, C., & Verrelst, J. (2020). DATimeS: A machine learning time series GUI toolbox for gap-filling and vegetation phenology trends detection. Environmental Modelling & Software, 104666.
How to cite: Amin, E., Belda, S., Pipia, L., Szantoi, Z., El Baroudy, A., Moreno, J., and Verrelst, J.: Crop phenology monitoring from Landsat 8 and Sentinel-2 green LAI time series at the Nile Delta, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-12437, https://doi.org/10.5194/egusphere-egu21-12437, 2021.
Abstract Climate change is a hot issue in the global scale. The some varieties of phenological phase of plants (trees, grasslands and crops et al.) can directly and objectively reflected climate change and Commonly, the response of plant phenology to climate change is sensitive, especially to climatic factors such as precipitation, temperature, soil characteristics in the growing environment, and sometimes can be considered as an indicator of climate change. Those meteorology and soil factors must be taken into account when we build phenological model so as to quantitatively study the relationship between climate change and plant phenology. Beside of those factors, the high frequency and multi-scale acquisition of phenological observation data is also the basis for phenological model researches. Since February of 2020, China Meteorological Administration (CMA) has established 25 vegetation ecological observation sites in Inner Mongolia autonomous region, Shaanxi, Hebei, Sichuan, Guangxi, Fujian and Anhui provinces. The automatic vegetation eco-meteorological observation instruments, whichi are composed of image sensor (digital camera), multispectral sensor, laser altimeter, point cloud laser radar and sound sensor, have been installed in the sites. They can provide so much products as image of plant community, normal difference vegetation index (NDVI), plant height, canopy height and animal sound at present. Of all these products, image data of plant community can be further retrieved to generate the greenness chromatic coordinate (Gcc) data, which can be widely applied into the phenological studies and the validations of satellite terrestrial vegetation products. After months of experimental operation, these equipments show the great ability to monitor the growth and development of terrestrial plants in China. This ability also lays a foundation for the establishment of the plant ecological observation network in China (China Vegetation Ecological Meteorological Observation Network).
KEYWORDS：Plant phenology, near-surface-based measurement, observation network
How to cite: Yang, D., Cui, S., Zhang, B., Wu, D., Lei, Y., Shen, C., Zhu, J., Wang, Y., and Wang, W.: Near-Ground-Based Optical Plant Phenology Measurements at China Ecological Meteorology Sites, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-12576, https://doi.org/10.5194/egusphere-egu21-12576, 2021.
In general, modeling phenological evolution represents a challenging task mainly because of time series gaps and noisy data, coming from different viewing and illumination geometries, cloud cover, seasonal snow and the interval needed to revisit and acquire data for the exact same location. For that reason, the use of reliable gap-filling fitting functions and smoothing filters is frequently required for retrievals at the highest feasible accuracy. Of specific interest to filling gaps in time series is the emergence of machine learning regression algorithms (MLRAs) which can serve as fitting functions. Among the multiple MLRA approaches currently available, the kernel-based methods developed in a Bayesian framework deserve special attention because of both being adaptive and providing associated uncertainty estimates, such as Gaussian Process Regression (GPR).
Recent studies demonstrated the effectiveness of GPR for gap-filling of biophysical parameter time series because the hyperparameters can be optimally set for each time series (one for each pixel in the area) with a single optimization procedure. The entire procedure of learning a GPR model only relies on appropriate selection of the type of kernel and the hyperparameters involved in the estimation of input data covariance. Despite its clear strategic advantage, the most important shortcomings of this technique are the (1) high computational cost and (2) memory requirements of their training, which grows cubically and quadratically with the number of model’s samples, respectively. This can become problematic in view of processing a large amount of data, such as in Sentinel-2 (S2) time series tiles. Hence, optimization strategies need to be developed on how to speed up the GPR processing while maintaining the superior performance in terms of accuracy.
To mitigate its computational burden and to address such shortcoming and repetitive procedure, we evaluated whether the GPR hyperparameters can be preoptimized over a reduced set of representative pixels and kept fixed over a more extended crop area. We used S2 LAI time series over an agricultural region in Castile and Leon (North-West Spain) and testing different functions for Covariance estimation such as exponential Kernel, Squared exponential kernel and matern kernel with parameter 3/2 or 5/2. The performance of image reconstructions was compared against the standard per-pixel GPR time series training process. Results showed that accuracies were on the same order (12% RMSE degradation) whereas processing time accelerated up to 90 times. Crop phenology indicators were also calculated and compared, revealing similar temporal patterns with differences in start and end of growing season of no more than five days. To the benefit of crop monitoring applications, all the gap-filling and phenology indicators retrieval techniques have been implemented into the freely downloadable GUI toolbox DATimeS (Decomposition and Analysis of Time Series Software - https://artmotoolbox.com/).
How to cite: Belda, S., Salinero, M., Amin, E., Pipia, L., Morcillo-Pallarés, P., and Verrelst, J.: Gaussian Process Regression hyperparameter optimization for image time series gap-filling of Earth observation data and crop monitoring, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-14322, https://doi.org/10.5194/egusphere-egu21-14322, 2021.
We are sorry, but presentations are only available for users who registered for the conference. Thank you.
We are sorry, but presentations are only available for users who registered for the conference. Thank you.