Phenological Transition Dates Research Articles

Quantifying City- and Street-Scale Urban Tree Phenology from Landsat-8, Sentinel-2, and PlanetScope Images: A Case Study in Downtown Beijing

Understanding the phenology of urban trees can help mitigate the heat island effect by strategically planting and managing trees to provide shade, reduce energy consumption, and improve urban microclimates. In this study, we carried out the first evaluation of high spatial resolution satellite images from Landsat-8, Sentinel-2, and PlanetScope images to quantify urban street tree phenology in downtown Beijing. The major research goals are to evaluate the consistency in pixel-level spring–summer growth period phenology and to investigate the capacity of high-resolution satellite observations to distinguish phenological transition dates of urban street trees. At the city scale, Landsat-8, Sentinel-2, and PlanetScope show similar temporal NDVI trends in general. The pixel-level analysis reveals that green-up date consistency is higher in areas with medium (NDVI > 0.5) to high (NDVI > 0.7) vegetation cover when the impacts of urban surfaces on vegetation reflectance are excluded. Similarly, maturity date consistency significantly increases in densely vegetated pixels with NDVI greater than 0.7. At the street scale, this study emphasizes the efficacy of NDVI time series derived from PlanetScope in quantifying the phenology of common street tree genera, including Poplars (Populus), Ginkgos (Ginkgo), Chinese Scholars (Styphnolobium), and Willows (Salix), in downtown Beijing to improve urban vegetation planning. Based on PlanetScope observations, we found that the four street tree genera have unique phenological patterns. Interestingly, we found that the trees along many major streets, where Chinese Scholars are the major tree genus, have later green-up dates than other areas in downtown Beijing. In conclusion, the three satellite observation datasets prove to be effective in monitoring street tree phenology during the spring–summer growth period in Beijing. PlanetScope is effective in monitoring tree phenology at the street scale; however, Landsat-8 may be affected by the mixture of land covers due to its relatively coarse spatial resolution.

Using Geostationary Satellite Observations to Improve the Monitoring of Vegetation Phenology

Geostationary satellite data enable frequent observations of the Earth’s surface, facilitating the rapid monitoring of land covers and changes. However, optical signals over vegetation, represented by the vegetation index (VI), exhibit an anisotropic effect due to the diurnal variation in the solar angle during data acquisition by geostationary satellites. This effect, typically characterized by the bi-directional reflectance distribution function (BRDF), can introduce uncertainties in vegetation monitoring and the estimation of phenological transition dates (PTDs). To address this, we investigated the diurnal variation in the normalized difference vegetation index (NDVI) with solar angles obtained from geostationary satellites since the image had fixed observation angles. By establishing a temporal conversion relationship between instantaneous NDVI and daily NDVI at the local solar noon (LSNVI), we successfully converted NDVIs obtained at any time during the day to LSNVI, increasing cloud-free observations of NDVI by 34%. Using different statistics of the time series vegetation index, including LSNVI, daily averaged NDVI (DAVI), and angular corrected NDVI (ACVI), we extracted PTD at five typical sites in China. The results showed a difference of up to 41.5 days in PTD estimation, with the highest accuracy achieved using LSNVI. The use of the proposed conversion approach, utilizing time series LSNVI, reduced the root mean square error (RMSE) of PTD estimation by 9 days compared with the use of actual LSNVI. In conclusion, this study highlights the importance of eliminating BRDF effects in geostationary satellite observations and demonstrates that the proposed angular normalization method can enhance the accuracy of time series NDVI in vegetation monitoring.

Comparison of change-based and shape-based data fusion methods in fine-resolution land surface phenology monitoring with Landsat and Sentinel-2 data

Fine-resolution land surface phenology (LSP) is urgently required for applications on agriculture management and vegetation-climate interaction, especially over heterogeneous areas, such as agricultural lands and fragmented forests. The critical challenge of fine-resolution LSP monitoring is how to reconstruct the spatiotemporal continuous vegetation index time series. To solve this problem, various data fusion methods have been devised; however, the comprehensive inter-comparison is lacking across different spatial heterogeneity, data quality, and vegetation types. We divide these methods into two main categories: the change-based methods fusing satellite observations with different spatiotemporal resolutions, and the shape-based methods fusing prior knowledge of shape models and satellite observations. We selected four methods to rebuilt two-band enhanced vegetation index (EVI2) series based on the harmonized Landsat and Sentinel-2 (HLS) data, including two change-based methods, namely the Spatial and temporal Adaptive Reflectance Fusion Model (STARFM), the Flexible Spatiotemporal DAta Fusion (FSDAF), and two shape-based methods, namely the Multiple-year Weighting Shape-Matching (MWSM), and the Spatiotemporal Shape-Matching Model (SSMM). Four phenological transition dates were extracted, evaluated with PhenoCam observations and the 500 m Visible Infrared Imaging Radiometer Suite (VIIRS) phenology product. The 30 m transition dates show more spatial details and reveal more apparent intra-class and inter-class phenology variation compared with 500 m product. The four transition dates of SSMM and FSDAF (R2>0.74, MAD<15 days) show better agreement with PhenoCam-derived dates. The performance difference between fusion methods over various application scenarios are then analyzed. Fusion results are more robust when temporal frequency is higher than 15 observations per year. The shape-based methods are less sensitive to temporal sampling irregularity than change-based methods. Both change-based methods and shape-based methods cannot perform well when the region is heterogeneous. Among different vegetation types, SSMM-like methods have the highest overall accuracy. The findings in this paper can provide references for regional and global fine-resolution phenology monitoring.

Drier August and colder September slow down the delaying trend of leaf senescence in herbaceous plants on the Qinghai–Tibetan Plateau

Plant phenological shifts on the Qinghai–Tibetan Plateau (QTP) have gained considerable attention over the last few decades. However, temporal changes in plant autumn phenology and the main driving factors remain uncertain. Most previous studies used satellite-derived phenological transition dates and climatic statistics during the preseason, which have relatively large uncertainties and may mask some important climate change characteristics at the intra-annual scale, thus affecting exploration of the underlying phenological change causes. This study collected 1685 phenological records at 27 ground stations on the QTP during 1983–2017. Temporal change trends and break points in leaf senescence date (LSD) of 23 herbaceous species were assessed using least squares regression, a meta-analysis procedure, and the Pettitt test. The main drivers and causes were investigated through correlation analysis and contribution calculation based on LSD observations and monthly climatic data. Results showed that, LSD of QTP herbaceous plants was significantly delayed at a rate of 4.45 days/decade during 1983–2017. Break points were concentrated during 1999–2003, with an overall mean in 2001. After 2001, the delay trend in LSD decreased, falling from 5.26 days/decade to 2.54 days/decade. Air temperature and precipitation were the most important climatic factors that showed closer and more extensive correlations with LSD and greater contributions to the inter-annual variations in LSD. August and September were the most critical period during which climatic factors had higher contributions to the LSD shifts. However, August was drier, with precipitation significantly decreasing and temperature increasing, and September was colder after 2001. Therefore, the declining trend in LSD may be attributed to the drier August and colder September. This study has not only provided reliable field evidence on temporal changes in autumn phenology on the QTP, but has also provided valuable insights into autumn phenological modelling and regional carbon cycling in alpine regions.

Improving the matching degree between remotely sensed phenological dates and physiological growing stages of soybean by a dynamic offset-adjustment strategy

Crop phenology provides crucial information for determining the appropriate timing of farm management practices and predicting crop yields. Satellite remote sensing has become a burgeoning tool for rapid phenological monitoring over wide spatial regions. However, there are significant timing gaps between the satellite-based phenological feature points and ground-observed physiological growing stages of the target. In this study, a dynamic offset-adjustment strategy that aims to improve the matching degree of the above two is proposed and tested with soybean across 16 states in the United States. A series of remotely sensed phenological transition dates that are characteristics of key growing stages of soybean were retrieved using MODIS time series data over the period 2000–2020 and the offset adjustments to the dates were identified by dynamically adjusting offset values till the minimum RMSE between the remote sensing-based and the ground-observed dates of physiological growing stages were obtained. The results indicated that the offset-adjustment strategy can significantly improve the alignments between remotely sensed phenological dates and field-based physiological growing stages of soybean in contrast to these without taking adjustment, with the average RMSE dropping by 58.58 %, 51.59 %, 31.15 %, 25.33 %, 24.67 % in the downturn, peak of season (POS), upturn, stabilization and recession dates, respectively. Among tested remotely sensed characteristics, the end of season (EOS) dates show the greatest alignment with its corresponding physiological growing stage, i.e., the dropping leaves stage. Comparison of the performance of the upturn date and start of season (SOS) in monitoring the date of the emerged stage indicates that the later one exhibits a better consistency with the ground-observed emerged stage after taking the adjustment, with the average RMSE dropping by 56.52 %. The proposed offset-adjustment strategy offers an approach for adjusting remotely sensed characteristics so to make them more consistent with the ground-observed crop physiological growing stages.

Methodological challenges and new perspectives of shifting vegetation phenology in eddy covariance data

While numerous studies report shifts in vegetation phenology, in this regard eddy covariance (EC) data, despite its continuous high-frequency observations, still requires further exploration. Furthermore, there is no general consensus on optimal methodologies for data smoothing and extracting phenological transition dates (PTDs). Here, we revisit existing methodologies and present new prospects to investigate phenological changes in gross primary productivity (GPP) from EC measurements. First, we present a smoothing technique of GPP time series through the derivative of its smoothed annual cumulative sum. Second, we calculate PTDs and their trends from a commonly used threshold method that identifies days with a fixed percentage of the annual maximum GPP. A systematic analysis is performed for various thresholds ranging from 0.1 to 0.7. Lastly, we examine the relation of PTDs trends to trends in GPP across the years on a weekly basis. Results from 47 EC sites with long time series (> 10 years) show that advancing trends in start of season (SOS) are strongest at lower thresholds but for the end of season (EOS) at higher thresholds. Moreover, the trends are variable at different thresholds for individual vegetation types and individual sites, outlining reasonable concerns on using a single threshold value. Relationship of trends in PTDs and weekly GPP reveal association of advanced SOS and delayed EOS to increase in immediate primary productivity, but not to the trends in overall seasonal productivity. Drawing on these analyses, we emphasise on abstaining from subjective choices and investigating relationship of PTDs trend to finer temporal trends of GPP. Our study examines existing methodological challenges and presents approaches that optimize the use of EC data in identifying vegetation phenological changes and their relation to carbon uptake.

A soil-air temperature model to determine the start of season phenology of deciduous forests

Forest ecosystems play a major role in sequestering atmospheric carbon dioxide, which can help offset the detrimental effects of anthropogenic carbon emissions. However, climate change has and will continue to affect the phenology of forest ecosystems’ carbon uptake, changing both the “carbon uptake transition date” - when forests shift from being a net carbon source to sink - and the “green-up date” reflecting the onset of bud burst. Previous studies have shown that a forest's carbon uptake transition date correlates to the date when soil temperature warms enough to surpass mean annual air temperature (soil-air temperature model). However, we still don't know if this simple relationship holds across different sites or over longer time periods. In this study, we explore the relationship between climate and both types of phenological transition dates using over 200 site years of data between 1997 and 2022. Using flux tower data from 18 sites across North America and Europe, we derive three potential carbon uptake transition dates corresponding to the dates when 10%, 25%, and 50% of seasonal net ecosystem exchange (NEE) amplitude is reached. Using PhenoCam data, we then derive three potential green-up dates corresponding to when 10%, 25%, and 50% of total seasonal green chromatic coordinate (GCC) is reached (the greenness model). We evaluate our model estimates using concordance coefficients, a metric of agreement between two measures, to determine which process, carbon uptake or budburst, is best predicted by the soil-air temperature model and to what extent. We find that variation in phenological relationships can be attributed to different regional and bioclimatic groups, highlighting potential biome-specific strengths and limitations of the soil-air temperature model. This model offers a simple approach to better understand phenological transitions and identify potential and limitations for a simple universal SOS prediction approach in deciduous forests.

Solar-induced chlorophyll fluorescence captures photosynthetic phenology better than traditional vegetation indices

Accurate characterization of plant phenology is of great importance for monitoring global carbon, water, and energy cycling. Remotely sensed satellite observations have been widely used to estimate land surface phenology across multiple spatial scales in the last three decades. Recent development on satellite solar-induced chlorophyll fluorescence (SIF) observations have opened an opportunity to monitor the seasonality of plant growth from the perspective of photosynthesis phenology. The SIF observations from the TROPOspheric Monitoring Instrument (TROPOMI) with high spatial resolution (up to 7 km × 3.5 km pixels) and near-daily global coverage provide unprecedented opportunity to observe photosynthetic and land surface phenology from space. However, the performance of TROPOMI SIF-derived phenology has not been systematically evaluated. In this study, we used flux tower gross primary productivity (GPP) and PhenoCam green chromatic coordinate (Gcc) data as the benchmark to verify phenology metrics derived from satellite observations. The phenology metrics including the start (SOS), end (EOS), length of growing season (LOS), and the peak of growing season (POS) were estimated from TROPOMI SIF, normalized difference vegetation index (NDVI), enhanced vegetation index (EVI), near‐infrared reflectance of vegetation (NIRv), and Global Vegetation Phenology product (MCD12Q2), and the latter four were obtained from the Moderate Resolution Imaging Spectroradiometer (MODIS) across six vegetation types over North America during the period of 2018–2020. We found that the overall agreements of SIF (R2 ranging from 0.30 to 0.63) against GPP-estimated phenology were stronger than NDVI (0.17–0.41), EVI (0.19–0.39), NIRv (0.23–0.62), and MCD12Q2 (0.19–0.48) derived phenological events. In reference to GPP and Gcc-estimated phenology, SIF also generally has the least error and bias compared to other satellite remote sensing-derived phenology metrics. No significant differences were found between SIF and GPP-derived phenology (P>0.05, two-tailed t-test). In addition, we found that the spatial distribution of SIF-derived phenology reflected the expected latitudinal patterns in phenology dates. SOS, EOS, LOS, and POS observed by MCD12Q2 appeared to be earlier, later, longer, and earlier than TROPOMI SIF-derived phenology, respectively. SIF-based phenological transition dates more closely tracked GPP-based phenology dates, indicating TROPOMI SIF could be a great measure to track photosynthesis seasonality and land surface phenology.

Using long-term data from a whole ecosystem warming experiment to identify best spring and autumn phenology models.

Predicting vegetation phenology in response to changing environmental factors is key in understanding feedbacks between the biosphere and the climate system. Experimental approaches extending the temperature range beyond historic climate variability provide a unique opportunity to identify model structures that are best suited to predicting phenological changes under future climate scenarios. Here, we model spring and autumn phenological transition dates obtained from digital repeat photography in a boreal Picea-Sphagnum bog in response to a gradient of whole ecosystem warming manipulations of up to +9°C, using five years of observational data. In spring, seven equally best-performing models for Larix utilized the accumulation of growing degree days as a common driver for temperature forcing. For Picea, the best two models were sequential models requiring winter chilling before spring forcing temperature is accumulated. In shrub, parallel models with chilling and forcing requirements occurring simultaneously were identified as the best models. Autumn models were substantially improved when a CO2 parameter was included. Overall, the combination of experimental manipulations and multiple years of observations combined with variation in weather provided the framework to rule out a large number of candidate models and to identify best spring and autumn models for each plant functional type.

Phenological assessment of transpiration: The stem-temp approach for determining start and end of season

Field-based assessment of transpiration phenology in boreal tree species is a significant challenge. Here we develop an objective approach that uses stem radius change and its correlation with sapwood temperature to determine the timing of phenological changes in transpiration in mixed evergreen species. We test the stem-temp approach using a five year stem-radius dataset from black spruce (Picea mariana) and jack pine (Pinus banksiana) trees in Saskatchewan (2016–2020). We further compare transpiration phenological transition dates from this approach with tower-based phenological assessment from green chromatic coordinate derived from phenocam images, eddy-covariance-derived evapotranspiration and carbon uptake, tower-based measurements of solar-induced chlorophyll fluorescence and snowmelt timing. The stem-temp approach identified the start and end of four key transpiration phenological phases: (i) the end of temperature-driven cycles indicating the start of biological activity, (ii) the onset of stem rehydration, (iii) the onset of transpiration, and (iv) the end of transpiration-driven cycles. The proposed method is thus useful for characterizing the timing of changes in transpiration phenology and provides information about distinct processes that cannot be assessed with canopy-level phenological measurements alone.

Assessing Canopy Phenological Variations and Gross Primary Productivity in A Savanna Ecosystem in Yuanjiang, Yunnan Province of Southwest China

Vegetation phenology is an important indicator of environmental change and strongly connected to forest ecosystem productivity change. This study aimed to analyse the pattern of phenological variations derived from digital imagery for the interpretation of ecosystem productivity. For 2014, 2015 and 2016, the seasonal phenological development of savanna was analysed by using towerbased imagery from a digital camera. The green excess index (GEI) was the best at representing the phenological transition dates (PTDs) and useful for investigating the gross primary production (GPP) in the savanna ecosystem. There was a significant correlation between the monthly pattern of the strength of green (Sgreen), green excess index (GEI) and vegetation contrast index (VCI) and GPP throughout the year. Additionally, the annual pattern of colour indices had significant relationship (p &lt; 0.05) with GPP but this was not seasonal. The air temperature (air T) and soil temperature (soil T) were strongly significantly correlated (p &lt; 0.001) with the start of growing season (SGS) and caused the advance in green-up and the timing of the start of the growing season in 2014 and 2016. The short growing season length (GSL) had an impact on the productivity. The colour indices from the digital camera images not only provided the phenological pattern of a forest canopy but also revealed the forest ecosystem productivity by showing the response to environmental factors. Our results indicate that daily continuous digital camera images might be useful for ecologists to use as a tool for future prediction of the long-term phenological modelling.

Monitoring nature's calendar from space: Emerging topics in land surface phenology and associated opportunities for science applications.

Vegetation phenology has been viewed as the nature's calendar and an integrative indicator of plant-climate interactions. The correct representation of vegetation phenology is important for models to accurately simulate the exchange of carbon, water, and energy between the vegetated land surface and the atmosphere. Remote sensing has advanced the monitoring of vegetation phenology by providing spatially and temporally continuous data that together with conventional ground observations offers a unique contribution to our knowledge about the environmental impact on ecosystems as well as the ecological adaptations and feedback to global climate change. Land surface phenology (LSP) is defined as the use of satellites to monitor seasonal dynamics in vegetated land surfaces and to estimate phenological transition dates. LSP, as an interdisciplinary subject among remote sensing, ecology, and biometeorology, has undergone rapid development over the past few decades. Recent advances in sensor technologies, as well as data fusion techniques, have enabled novel phenology retrieval algorithms that refine phenology details at even higher spatiotemporal resolutions, providing new insights into ecosystem dynamics. As such, here we summarize the recent advances in LSP and the associated opportunities for science applications. We focus on the remaining challenges, promising techniques, and emerging topics that together we believe will truly form the very frontier of the global LSP research field.

Potential of using surface temperature data to benchmark Sentinel-2-based forest phenometrics in boreal Finland

Key messageWe present a new approach to calibrate timings of phenological events from satellite data (e.g., Sentinel-2 MSI data) with readily available surface temperature data. The new approach improves the estimation of growing season length in boreal forests.ContextSatellite data is used to calibrate phenology models employed in land surface model components of climate models. However, realistic quantification of forest phenological transitions, such as the greenup and senescence, across large spatial scales remains challenging due to the lack of sufficient ground validation data representative of both forest tree canopy and forest understory species compositions.AimsThe aim of this study was to develop a new approach to benchmark boreal forest land surface phenology obtained from Sentinel-2 (S2) against surface temperature data.MethodsWe computed S2 phenological transition dates and compared them to ground reference data on temperature from a network of meteorological stations across Finland (60–70N°).ResultsOur results showed that applying standard phenometrics directly to S2 data to estimate the growing season length in boreal forests may lead to clear biases in all species groups.ConclusionOur approach to use temperature data to calibrate boreal forest phenometrics allows flexible application across spatial scales (i.e., point or grid) and different satellite sensors. It can be combined with any vegetation land cover product to provide a link between surface temperature data and forest seasonal reflectance properties.

A Robust Hybrid Deep Learning Model for Spatiotemporal Image Fusion

Dense time-series remote sensing data with detailed spatial information are highly desired for the monitoring of dynamic earth systems. Due to the sensor tradeoff, most remote sensing systems cannot provide images with both high spatial and temporal resolutions. Spatiotemporal image fusion models provide a feasible solution to generate such a type of satellite imagery, yet existing fusion methods are limited in predicting rapid and/or transient phenological changes. Additionally, a systematic approach to assessing and understanding how varying levels of temporal phenological changes affect fusion results is lacking in spatiotemporal fusion research. The objective of this study is to develop an innovative hybrid deep learning model that can effectively and robustly fuse the satellite imagery of various spatial and temporal resolutions. The proposed model integrates two types of network models: super-resolution convolutional neural network (SRCNN) and long short-term memory (LSTM). SRCNN can enhance the coarse images by restoring degraded spatial details, while LSTM can learn and extract the temporal changing patterns from the time-series images. To systematically assess the effects of varying levels of phenological changes, we identify image phenological transition dates and design three temporal phenological change scenarios representing rapid, moderate, and minimal phenological changes. The hybrid deep learning model, alongside three benchmark fusion models, is assessed in different scenarios of phenological changes. Results indicate the hybrid deep learning model yields significantly better results when rapid or moderate phenological changes are present. It holds great potential in generating high-quality time-series datasets of both high spatial and temporal resolutions, which can further benefit terrestrial system dynamic studies. The innovative approach to understanding phenological changes’ effect will help us better comprehend the strengths and weaknesses of current and future fusion models.

Hybrid phenology matching model for robust crop phenological retrieval

Crop phenology regulates seasonal agroecosystem carbon, water, and energy exchanges, and is a key component in empirical and process-based crop models for simulating biogeochemical cycles of farmlands, assessing gross and net primary production, and forecasting the crop yield. The advances in phenology matching models provide a feasible means to monitor crop phenological progress using remote sensing observations, with a priori information of reference shapes and reference phenological transition dates. Yet the underlying geometrical scaling assumption of models, together with the challenge in defining phenological references, hinders the applicability of phenology matching in crop phenological studies. The objective of this study is to develop a novel hybrid phenology matching model to robustly retrieve a diverse spectrum of crop phenological stages using satellite time series. The devised hybrid model leverages the complementary strengths of phenometric extraction methods and phenology matching models. It relaxes the geometrical scaling assumption and can characterize key phenological stages of crop cycles, ranging from farming practice-relevant stages (e.g., planted and harvested) to crop development stages (e.g., emerged and mature). To systematically evaluate the influence of phenological references on phenology matching, four representative phenological reference scenarios under varying levels of phenological calibrations in terms of time and space are further designed with publicly accessible phenological information. The results indicate that the hybrid phenology matching model can achieve high accuracies for estimating corn and soybean phenological growth stages in Illinois, particularly with the year- and region-adjusted phenological reference (R-squared higher than 0.9 and RMSE less than 5 days for most phenological stages). The inter-annual and regional phenological patterns characterized by the hybrid model correspond well with those in the crop progress reports (CPRs) from the USDA National Agricultural Statistics Service (NASS). Compared to the benchmark phenology matching model, the hybrid model is more robust to the decreasing levels of phenological reference calibrations, and is particularly advantageous in retrieving crop early phenological stages (e.g., planted and emerged stages) when the phenological reference information is limited. This innovative hybrid phenology matching model, together with CPR-enabled phenological reference calibrations, holds considerable promise in revealing spatio-temporal patterns of crop phenology over extended geographical regions.

Assessing the inter-annual variability of vegetation phenological events observed from satellite vegetation index time series in dryland sites

Tracking dryland vegetation phenology under a changing climate is of great concern because dryland ecosystems have broad spatial coverage and are important drivers of global carbon cycles. However, dryland ecosystems often consist of two or more different vegetation types with divergent phenological patterns and are characterized by relatively low vegetation greenup signal compared to noise present in satellite imagery. Thus, accurately characterizing dryland phenology – including the start, peak, and end of season (SOS, POS, and EOS, respectively) – across large temporal and spatial scales with satellite vegetation indices (VIs) has proven challenging. Working across six dryland flux tower sites in the United States from 2003 to 2014, we asked: 1) How well do satellite VIs explain GPP? 2) How accurately can satellite VIs characterize the number and timing of phenological transition dates? and 3) Why does this accuracy vary across VIs? To address these questions, we used daily GPP data from the FLUXNET2015 dataset. We computed four daily VIs from the Moderate Resolution Imaging Spectrometer (MODIS) sensor including two greenness indices (i.e., NDVI and EVI), and two water-related indices (i.e., Land Surface Water Index (LSWI) and NDVI-LSWI difference). We derived phenological transition dates from spline smoothed VI and GPP time series. We found that all major phenological patterns apparent in GPP time series were present in VI data across sites and VIs. Among the four VIs, EVI explained the most variance in daily GPP (R2 = 0.65). Between 88 and 90% of phenological events found in GPP time series – especially recurring, large magnitude events but also some sporadic, small ephemeral events – were detected from MODIS VI data. The rate of omission, when an event was detected from GPP but not from VI data, was similar across VIs (10–12% of all GPP events). Commissions, when an event was detected from VI but not from GPP data, varied more widely from 18% for NDVI to 30% for LSWI. Transition date accuracy as measured by mean absolute difference (MAD) between GPP- and VI-derived SOS and POS ranged from 13 to 15 days across VIs but was lower for EOS (17–26 days). EOS from LSWI most accurately characterized GPP-derived EOS. This may be because LSWI is sensitive to leaf water content, which declines in certain dryland vegetation even when leaves remain green. Our results highlight the potential for satellite VI-derived metrics to accurately track spatio-temporal variation in the phenological event occurrence and timing in dryland ecosystems.

Continuous observations of forest canopy structure using low-cost digital camera traps

Assessing forest canopy dynamics is crucial for understanding the response of vegetation to environmental variability and change. While digital repeat photography is gaining increased attention for obtaining field phenology observations, colour indices derived from this method are often affected by leaf colour and actual canopy structure, complicating the physical interpretation of results. In addition, repeated photography requires power, storage capacity and remote data transfer, which are often limited in forest conditions. As an alternative, we tested a simple, cheap and fast solution to derive daily canopy structure observation from digital camera traps (CTs). Formerly deployed for wildlife monitoring, CTs are low-cost digital cameras designed for outdoor conditions and have low battery consumption, enable repeat acquisition, and often feature remote data transfer protocols.The trial was performed in a deciduous oak stand, where continuous images were acquired over a 1-year period using the time-lapse feature of the CT. Daily time series of canopy structure attributes were derived from the collected images using simple and automated procedures. Results were validated against reference manual cover photography measurements. The daily time series of foliage cover and leaf area index were then used to derive phenological transition dates, which were compared against phenological observations obtained from satellite Sentinel-2 data. Results indicated that field and satellite data provided comparable accuracy in determining the start of season (SOS). Larger discrepancies were found in determining the end of season (EOS), which can be attributed to the low number of good quality autumn images available from the satellite data. We concluded that CT is a robust method, which is ideally suited for routine, continuous field monitoring of canopy attributes and phenology. While this method can be used for evaluating remote sensing observations, the combination of CTs with satellite data holds great potential for greater spatiotemporal coverage, from field to landscape scales.

Trail camera networks provide insights into satellite-derived phenology for ecological studies

Repeat digital photography at or near ground-level is a proven and efficient approach for tracking plant phenology. Here, we explored the potential to monitor phenology using the Snapshot Wisconsin (SW) trail camera network, a citizen science program. Using three curve-fitting methods for characterizing phenological transition dates, we assessed the phenological offset between understory vegetation and the overstory canopy in the trailcam observations and compared variations in derived phenology over the different spatial scales represented by trailcams (~20–50 m), Harmonized Landsat and Sentinel-2 (HLS, 30 m), and Moderate Resolution Imaging Spectroradiometer (MODIS, 500 m). Our results showed that the apparent phenological offset between understory and overstory vegetation differed among forest types: in broadleaf deciduous forests, understory vegetation had an earlier start-of-spring (SOS) and later end-of-autumn (EOA) than the overstory canopy; in mixed forests, the understory showed an earlier SOS than the overstory, but no significant difference in EOA; in evergreen conifer forests, neither SOS nor EOA differed significantly between the understory and overstory. We found moderate correlations (0.25 ≤ r ≤ 0.57) between trailcam- and satellite-derived phenological dates. Moreover, those derived dates varied significantly among the applied curve-fitting methods: total growing season length (from SOS to EOA) could be 19 days longer for a threshold-based method than for a logistic curve-fitting method (our reference model), but 17 days shorter than the logistic method when using a piecewise-continuous method based on fitted sine curves. Despite the spatial limitations of trailcams for characterizing phenology on landscape and regional scales, trailcam networks have considerable potential for informing local phenological studies and disentangling the many drivers of phenology that can remain undetected from the satellite perspective.

Remote sensing phenological monitoring framework to characterize corn and soybean physiological growing stages

The phenological dynamics of crops reflect the response and feedback of agricultural systems to climate and environmental constraints, and have significant controls on carbon and nutrient cycling across the globe. Remote monitoring of crop phenological dynamics in a consistent and systematic manner is vitally crucial for optimizing the farm management activities and evaluating the agricultural resilience to extreme weather conditions and future climate change. Yet our ability to retrieve crop growing stages with satellite time series is limited. The remotely sensed phenological transition dates may not be characteristic of crop physiological growing stages. The objective of this study is to develop a remote sensing phenological monitoring framework that can reconcile satellite-based phenological measures with ground-based crop growing observations, with corn and soybean in Illinois as a case study. The framework comprises three key components: time series phenological pre-processing, time series phenological modeling, and time series phenological characterization. As an exploratory prototype, the framework retrieved a total of 56 phenological transition dates that were subsequently evaluated with the district-level ground phenological observations. The results indicated that the devised framework can adequately retrieve a wide range of physiological growing stages for corn and soybean in Illinois, with R square greater than 0.6 and RMSE less than 1 week for most stages. The devised framework largely extends the limited satellite phenological measures to a range of phenological transition dates that are characteristic of essential crop growing stages. It paves the way for formulating standard crop phenological monitoring protocols via remote sensing. The wealth of retrieved phenological characteristics open up unique opportunities to enhance our understanding of the complex mechanisms underlying the crop growth in response to varying environmental stresses, and to make more adaptive farm management strategies towards sustained agricultural development.

Understanding the continuous phenological development at daily time step with a Bayesian hierarchical space-time model: impacts of climate change and extreme weather events

The impacts of climate change and extreme weather events (e.g. frost-, heat-, drought-, and heavy rainfall events) on the continuous phenological development over the entire seasonal cycle remained poorly understood. Previous studies mainly focused on modeling key phenological transition dates (e.g. discrete timing of spring bud-break and fall senescence) based on aggregated climate variables (e.g. mean temperature, growing-degree days). We developed and evaluated a Bayesian Hierarchical Space-Time model for Land Surface Phenology (BHST-LSP) to synthesize remotely sensed vegetation greenness with climate covariates at a daily temporal scale from 1981 to 2014 across the entire conterminous United States. The BHST-LSP model incorporated both temporal and spatial information and exhibited high predictive power in simulating daily phenological development with an overall out-of-sample R2 of 0.80 ± 0.17 and 0.72 ± 0.20 for spring and fall phenology, respectively. The overall out-of-sample normalized root mean square errors were 9.3% ± 6.1% and 9.9% ± 5.2% between the observed and predicted vegetation greenness for spring and fall phenology, respectively. We found that a fast increase of temperature can accelerate the speed of spring green-up while a slow decrease of temperature can lead to a decelerated fall brown-down. Increasing accumulated precipitation can benefit daily phenological development over an entire growing season, while extreme rainfall events can have the opposite effects. More frequent frost events could slow spring leaf expansion and accelerate fall leaf senescence. Impacts of extreme heat events were complex and depended on water availability. Cropland in the Midwest as well as evergreen needleleaf forest along the coastal regions showed relatively strong resistance to drought events compared to other land cover types. The BHST-LSP model can be used to forecast vegetation phenology given future climate projection, thus providing valuable information for adopting climate change adaptation and mitigation measures.