Elsevier

Science of The Total Environment

Volume 644, 10 December 2018, Pages 439-451
Science of The Total Environment

Seasonal patterns of canopy photosynthesis captured by remotely sensed sun-induced fluorescence and vegetation indexes in mid-to-high latitude forests: A cross-platform comparison

https://doi.org/10.1016/j.scitotenv.2018.06.269Get rights and content

Highlights

  • Phenological metrics from 2 SIF sets and 4 vegetation indexes were intercompared.

  • Remotely sensed SIF were highly correlated with GPP in mid-to-high latitude forests.

  • The SIF-GPP relationships can be generally considered linear at 16-day scale.

  • PI and NDVI provided reliable predictions of start of seasons among MODIS indexes.

  • Limitations remained for OCO-2 SIF to extract photosynthesis phenology at site-level.

Abstract

Characterized by the noticeable seasonal patterns of canopy photosynthesis, mid-to-high latitude forests are sensitive to climate change and crucial for understanding the global carbon cycle. To monitor the seasonal cycle of the canopy photosynthesis from space, several remotely sensed indexes, such as normalized difference vegetation index (NDVI), enhanced vegetation index (EVI), and leaf area index (LAI) have been implemented within the past decades. Recently, satellite-derived sun-induced fluorescence (SIF) has shown great potential of providing retrievals that are more related to photosynthesis process. However, the potentials of different canopy measurements have not been thoroughly assessed in the context of recent advances of new satellites and proposals of improved indexes. At 15 forested sites, we present a cross-platform intercomparison of one emerging remote sensing based index of phenology index (PI) and two SIF datasets against the conventional indexes such as NDVI, EVI, and LAI to capture the seasonal cycles of canopy photosynthesis. NDVI, EVI, LAI, and PI were calculated from Moderate Resolution Imaging Spectroradiometer (MODIS) measurements, while SIF were evaluated from Global Ozone Monitoring Experiment-2 (GOME-2) and Orbiting Carbon Observatory-2 (OCO-2) observations. Results indicated that GOME-2 SIF was highly correlated with gross primary production (GPP) and absorbed photosynthetically active radiation during the growing seasons. The SIF-GPP relationship can generally be considered linear at the 16-day scale. Key phenological metrics such as start of the seasons and end of the seasons captured by SIF from GOME-2 and OCO-2 matched closely with photosynthesis phenology as inferred by GPP. However, the applications of OCO-2 SIF for phenological studies may be limited only for a small range of sites (at site-level) due to a limited spatial sampling. Among the MODIS estimations, PI and NDVI provided most reliable predictions of start of growing seasons, while no indexes accurately captured the end of growing seasons.

Graphical abstract

The original EC GPP and OCO-2 SIF with the fitted seasonal cycles and predicted SOS and EOS at six forested sites.

Unlabelled Image
  1. Download : Download high-res image (131KB)
  2. Download : Download full-size image

Introduction

Terrestrial ecosystems play an important role in regulating regional and global climate (Burrows et al., 2011). Mid-to-high latitude forests, especially the boreal forests, are substantial contributors to carbon fluxes (Beer et al., 2010; Rolleston, 1996). As plants in these regions are expected to experience the greatest warming among forest biomes, they are deemed to react and respond sensitively to climate change and variability (Keenan et al., 2014). In recent years, with the developments of networks of flux measurements and advances of remote sensing based models, the monitoring of the physiological processes such as photosynthesis of mid-to-high latitudes has become generally possible.

Mid-to-high latitude forests are showing noticeable seasonal cycles of canopy photosynthesis. These life cycle events are sensitive indicators of the biosphere's response to climate changes through contributions to the global carbon, energy and water cycles (Buitenwerf et al., 2015; Peñuelas et al., 2009). Understanding the changes of these cycles as well as the underlying mechanisms are of significance for predicting future changes of climate and the global carbon cycle. Recent in-situ and remote sensing based studies have shown that the warming climate has triggered lengthier growing seasons in northern hemisphere regions (Cleland et al., 2007; Viña et al., 2016; Wang et al., 2015). Remote sensing based approaches to estimate phenological metrics (e.g., the start and end of growing seasons) were mainly based on reflectance-calculated vegetation indexes (VIs), such as normalized difference vegetation index (NDVI), enhanced vegetation index (EVI) and leaf area index (LAI) retrieved using these VIs (Helman, 2018). These indexes have been applied to regional and global studies, especially for the regions without long-term ground observations (Gonsamo and Chen, 2016). Fundamentally, yet, those VIs cannot provide us a direct understanding of physiological processes so that can be hard to be perfectly applied to modelling frameworks. At the same time, several recent studies found that performances of VIs are significantly hindered by snow cover and soil moisture in high-latitude regions (D'Odorico et al., 2015; Peng et al., 2017; Wu et al., 2017). Several improved indexes including phenology index (PI) that aimed at the match between remotely sensed and ground observed seasonal cycles of canopy photosynthesis have been proposed in recent years (Gonsamo et al., 2012a). The PI combines NDVI and Normalized Difference Infrared Index (NDII) aiming to decouple the seasonality of the green vegetation component from the background one because green-up co-occurs with snow melt (Delbart et al., 2005; Gonsamo et al., 2012a). Yet, the biological recovery and dormancy for evergreen forests are still extremely difficult to identify during the transition period when the greenness signal of the vegetation is weak or does not necessarily correspond with the shifts of photosynthesis (Wong and Gamon, 2015).

Fortunately, recent advances of atmospheric measurements has made it possible to retrieve an alternative indicator that is more related to the photosynthesis processes: sun-induced fluorescence (SIF). Chlorophyll pigments absorb photons to power photosynthesis, with some of the photons are re-emitted at longer wavelengths as chlorophyll fluorescence (Baker, 2008). The re-emitted SIF has been successfully related to downward carbon flux, i.e., carbon uptake by the vegetation. This provides a promising way in estimating photosynthesis through SIF. Global SIF datasets using space-borne spectroscopy from satellites became available past few years (Frankenberg et al., 2011; Frankenberg et al., 2014; Guanter et al., 2013; Guanter et al., 2014; Joiner et al., 2013; Joiner et al., 2016; Köhler et al., 2015). Despite the complex processes underlying the relationship between SIF and gross primary production (GPP), it has been reported the satellite-retrieved SIF was highly correlated with GPP estimated based on eddy covariance (EC) flux towers (van der Tol et al., 2014; Verma et al., 2017; Yang et al., 2017; Yang et al., 2015; Zhang et al., 2016b). Their relationship appears to reflect the level of absorbed photosynthetically active radiation (APAR) with additional information of light use efficiency (LUE). Based on >50 EC towers, Joiner et al. (2014) found that the Global Ozone Monitoring Experiment-2 (GOME-2) SIF retrieved phenological metrics matched closely with that of EC-based estimations, although the footprints of GOME-2 (40 km by 80 km) were significantly larger than most EC sites. Walther et al. (2016) found that GOME-2 SIF decoupled growing seasons can be up to 6 weeks longer than that captured by EVI. Jeong et al. (2017) evaluated remotely sensed SIF and NDVI of several platforms and proposed that the continued measurements of SIF and NDVI would help us to understand the seasonal variations of vegetation photosynthesis and greenness. However, the coarse spatial representativeness of previous atmospheric measurements (~40 km by 80 km or coarser) makes it difficult to compare with ground-based canopy measurements (Chen et al., 2012; Joiner et al., 2014; Zhang et al., 2016b). Very recently, Orbiting Carbon Observatory 2 (OCO-2) has shown renewed promises of provinding satellite-derived SIF with the improved spatial representativeness at around 1.3 km by 2.25 km (Frankenberg et al., 2014). The footprints of OCO-2 that match the spatial representativeness of most EC towers enables it to produce better results (Lu et al., 2018; Verma et al., 2017). The emerging observations from OCO-2, however, have rarely been applied in phenological studies (Köhler et al., 2018).

In most physiological models, VIs and/or LAI were used to decouple the seasonal cycles of processes such as photosynthesis (Wang et al., 2016). To constrain the uncertainties of current models regarding the estimations of productivity, the use of high-resolution and global retrieval of SIF might further improve the accuracy. In the context of extreme events including prolonged droughts in recent decades, it was vital to comprehensively investigate the usability of SIF in monitoring canopy photosynthesis, including the key phenological metrics, compared with conventional VIs and LAI (Dahlin et al., 2015; Melaas et al., 2016; Zipper et al., 2016). At the same time, although several models that can estimate GPP globally with VIs have been proposed, the advantages of SIF can potentially be exploited to improve their performances especially at high-latitude regions (Jeong et al., 2017; Luus et al., 2017; Luus and Lin, 2015). However, the advantages of SIF compared against conventional VIs have not been comprehensively accessed in the context that there are advances of new satellites capable of monitoring SIF at relatively high spatial resolutions and proposal of improved VIs with different theories (Sun et al., 2017). In this study, our primary objective was to evaluate and compare the seasonal cycles of several remotely sensed canopy measurements across mid-to-high latitude forests with a focus on evergreen needleleaf forests (ENF), deciduous broadleaf forests (DBF) and mixed forests (MF). An additional objective was to focus on phenological transition dates derived from different platforms, which are indicators directly related to the carbon budgets of terrestrial ecosystems.

Section snippets

EC estimated canopy properties

We conducted this study at 15 EC sites in North America and Europe where relatively homogeneous landscapes exist around the flux towers. In total, there are 103 site-years with a mean of 6.9 years of observations for each site. These sites represent three main forest biomes in mid-to-high latitude forests such as ENF, DBF andMF (Fig. 1 and Table 1). The selection of EC sites was based on an assumption of threshold (60%) of International Geosphere-Biosphere Program (IGBP) classifications (

Relationship between satellite-derived SIF and EC towers based GPP

We used the OCO-2 SIF to explore the relationship of satellite-derived SIF and GPP at instantaneous scale at US-PFa and US-WCr first (Fig. 2). The EC instruments of US-PFa and US-WCr were located in a mixed forest and a deciduous forest in northern Wisconsin, USA, respectively. From late 2014 to 2017, there were 21 times of observations from OCO-2 at US-PFA and 20 times of observations at US-WCr. At both sites, the correlations of linear and hyperbolic models were statistically significant

SIF-GPP relationship

In this study, we focused on 15 EC sites in mid-to-high latitude forests to examine the seasonal trajectories of satellite-derived VIs, LAI, and SIF, as well as their relationships with EC-based estimations of canopy photosynthesis. An additional objective was to explore the capacities of five remote sensing based measurements to track the key seasonal metrics in photosynthesis.

Despite the potentials of SIF to estimate GPP at various spatial and temporal scales, the SIF-GPP relationship can be

Conclusions and outlooks

Our results added additional endorsements for the applications of satellite-derived SIF in phenological studies in forest biomes. In 15 mid-to-high latitude forests in North America and Europe, the seasonal trajectories of GOME-2 datasets were significantly correlated with GPPEC with R2 values ranged from 0.52 to 0.74 with the linear model, while that ranged from 0.54 to 0.74 with the hyperbolic model. At the same time, the start and end of growing seasons estimated by GOME-2 and OCO-2 SIF

Acknowledgments

This work was supported by the Chinese Arctic and Antarctic Administration, National Natural Science Foundation of China, the Chinese Polar Environment Comprehensive Investigation, Assessment Program (Grant No. 41676176 and No. 41676182). This work was also supported by the Fundamental Research Funds for the Central Universities (Grant No. 312231103). We acknowledge Dr. Köhler P. for sharing the satellite-derived SIF, Dr. Tramontana G. and Prof. M. Jung for providing the upscaled model GPP. The

References (88)

  • L. Guanter

    Using field spectroscopy to assess the potential of statistical approaches for the retrieval of sun-induced chlorophyll fluorescence from ground and space

    Remote Sens. Environ.

    (2013)
  • D. Helman

    Land surface phenology: what do we really ‘see’ from space?

    Sci. Total Environ.

    (2018)
  • A. Huete

    Overview of the radiometric and biophysical performance of the MODIS vegetation indices

    Remote Sens. Environ.

    (2002)
  • S.J. Jeong

    Application of satellite solar-induced chlorophyll fluorescence to understanding large-scale variations in vegetation phenology and function over northern high latitude forests

    Remote Sens. Environ.

    (2017)
  • C. Jin

    Effects of in-situ and reanalysis climate data on estimation of cropland gross primary production using the Vegetation Photosynthesis Model

    Agric. For. Meteorol.

    (2015)
  • J. Joiner

    The seasonal cycle of satellite chlorophyll fluorescence observations and its relationship to vegetation phenology and ecosystem atmosphere carbon exchange

    Remote Sens. Environ.

    (2014)
  • P. Köhler et al.

    Assessing the potential of sun-induced fluorescence and the canopy scattering coefficient to track large-scale vegetation dynamics in Amazon forests

    Remote Sens. Environ.

    (2018)
  • X. Li et al.

    Chlorophyll fluorescence observed by OCO-2 is strongly related to gross primary productivity estimated from flux towers in temperate forests

    Remote Sens. Environ.

    (2018)
  • Z. Liu

    Improved modeling of gross primary production from a better representation of photosynthetic components in vegetation canopy

    Agric. For. Meteorol.

    (2017)
  • X. Lu et al.

    Opportunities and challenges of applications of satellite-derived sun-induced fluorescence at relatively high spatial resolution

    Sci. Total Environ.

    (2018)
  • X. Luo et al.

    Modeling and predicting spring land surface phenology of the deciduous broadleaf forest in northern China

    Agric. For. Meteorol.

    (2014)
  • J.H. Mccaughey et al.

    Carbon dioxide and energy fluxes from a boreal mixedwood forest ecosystem in Ontario, Canada

    Agric. For. Meteorol.

    (2006)
  • R.B. Myneni

    Global products of vegetation leaf area and fraction absorbed PAR from year one of MODIS data

    Remote Sens. Environ.

    (2002)
  • T. Nakai

    Characteristics of evapotranspiration from a permafrost black spruce forest in interior Alaska

    Polar Sci.

    (2013)
  • D. Peng

    Spring green-up phenology products derived from MODIS NDVI and EVI: intercomparison, interpretation and validation using National Phenology Network and AmeriFlux observations

    Ecol. Indic.

    (2017)
  • D.A. Sims

    A new model of gross primary productivity for North American ecosystems based solely on the enhanced vegetation index and land surface temperature from MODIS

    Remote Sens. Environ.

    (2008)
  • Y. Sun

    Overview of Solar-Induced chlorophyll Fluorescence (SIF) from the Orbiting Carbon Observatory-2: retrieval, cross-mission comparison, and global monitoring for GPP

    Remote Sens. Environ.

    (2018)
  • D.P. Turner

    Scaling Gross Primary Production (GPP) over boreal and deciduous forest landscapes in support of MODIS GPP product validation

    Remote Sens. Environ.

    (2003)
  • A. Viña et al.

    Land surface phenology as an indicator of biodiversity patterns

    Ecol. Indic.

    (2016)
  • C. Wang et al.

    Temperature sensitivity of spring vegetation phenology correlates to within-spring warming speed over the Northern Hemisphere

    Ecol. Indic.

    (2015)
  • C. Wu

    Land surface phenology derived from normalized difference vegetation index (NDVI) at global FLUXNET sites

    Agric. For. Meteorol.

    (2017)
  • X. Xiao

    Satellite-based modeling of gross primary production in an evergreen needleleaf forest

    Remote Sens. Environ.

    (2004)
  • X. Xiao

    Modeling gross primary production of temperate deciduous broadleaf forest using satellite images and climate data

    Remote Sens. Environ.

    (2004)
  • X. Yang

    Seasonal variability of multiple leaf traits captured by leaf spectroscopy at two temperate deciduous forests

    Remote Sens. Environ.

    (2016)
  • Y. Zhang

    Model-based analysis of the relationship between sun-induced chlorophyll fluorescence and gross primary production for remote sensing applications

    Remote Sens. Environ.

    (2016)
  • Y. Zhang

    Consistency between sun-induced chlorophyll fluorescence and gross primary production of vegetation in North America

    Remote Sens. Environ.

    (2016)
  • S.H. Alemohammad

    Water, Energy, and Carbon with Artificial Neural Networks (WECANN): a statistically based estimate of global surface turbulent fluxes and gross primary productivity using solar-induced fluorescence

    Biogeosciences

    (2017)
  • A. Anav

    The role of plant phenology in stomatal ozone flux modelling

    Glob. Chang. Biol.

    (2017)
  • K.M. Attard et al.

    Benthic primary production and mineralization in a High Arctic fjord: in situ assessments by aquatic eddy covariance

    Mar. Ecol. Prog. Ser.

    (2016)
  • G. Badgley et al.

    Canopy near-infrared reflectance and terrestrial photosynthesis

    Sci. Adv.

    (2017)
  • N.R. Baker

    Chlorophyll fluorescence: a probe of photosynthesis in vivo

    Annu. Rev. Plant Biol.

    (2008)
  • D. Baldocchi

    FLUXNET: a new tool to study the temporal and spatial variability of ecosystem-scale carbon dioxide, water vapor, and energy flux densities

    Bull. Am. Meteorol. Soc.

    (2001)
  • A.G. Barr

    Comparing the carbon budgets of boreal and temperate deciduous forest stands

    Can. J. For. Res.

    (2002)
  • C. Beer

    Terrestrial gross carbon dioxide uptake: global distribution and covariation with climate

    Science

    (2010)
  • Cited by (20)

    • How well do recently reconstructed solar-induced fluorescence datasets model gross primary productivity?

      2022, Remote Sensing of Environment
      Citation Excerpt :

      The comparatively high spatial resolution (250 m – 0.05°), temporal resolution (daily), and spectral resolution (> 6 bands for vegetation monitoring) make MODIS the most feasible option to gap-fill and downscale satellite SIF data at different temporal resolutions (4-day, 8-day, 16-day, or monthly) and different spatial resolutions (0.05° × 0.05° and 0.5° × 0.5° degrees) (Zhang et al., 2018a). These SIFr datasets have recently been extensively employed for capturing interannual variability of global and regional GPP (Bai et al., 2021; Chen et al., 2021b; Li and Xiao, 2019b, 2020; Sloat et al., 2021; Tagesson et al., 2021), impact of drought and moisture stress on vegetation (He et al., 2020; Li et al., 2020; Lopes Ribeiro et al., 2021; Xu et al., 2021; Zhang et al., 2021) and most likely would continue to be used for understanding global and regional GPP dynamics. However, it is not clear how these different SIFr datasets are better in predicting GPP than the widely used vegetation indices - NDVI (normalized difference vegetation index) and EVI (enhanced vegetation index), and which one performs the best and how to get the most out of these exciting datasets.

    • Recent trends and advances in hyperspectral imaging techniques to estimate solar induced fluorescence for plant phenotyping

      2022, Ecological Indicators
      Citation Excerpt :

      The results reveal that optimal results can be approximated when SIF and PRI are considered for estimating the assimilation rate of CO2. ( Xinchen Lu et al., 2018) reveals the pattern of seasonal variations in canopy photosynthesis using a hyperbolic model (Non-Linear). Several studies show that the SIF-GPP relationship is not linear, so the LUE can be correlated with PAR using a hyperbolic function.

    • Land surface phenology detections from multi-source remote sensing indices capturing canopy photosynthesis phenology across major land cover types in the Northern Hemisphere

      2022, Ecological Indicators
      Citation Excerpt :

      The mounting evidences indicated that SIF was closely related to gross primary production (GPP) according to a number of studies from different regions and satellites (Li et al., 2018; Sun et al., 2017). GPP based phenology as a key indicator of vegetation actual photosynthesis activity was treated as real carbon phenology to verify the accuracy of remotely sensed phenology (Lu et al., 2018a; Wang et al., 2019b; Wu et al., 2013). The comparison of land surface phenology obtained from SIF and GPP indicated that SIF can track the seasonal dynamics of vegetation development well for forests (Jeong et al., 2017).

    View all citing articles on Scopus
    View full text