Search

Although rice cultivation is one of the most important agricultural sources of methane (CH4) and contributes ∼8% of total global anthropogenic emissions, large discrepancies remain among estimates of global CH4 emissions from rice cultivation (ranging from 18 to 115 Tg CH4 yr−1) due to a lack of observational constraints. The spatial distribution of paddy-rice emissions has been assessed at regional-to-global scales by bottom-up inventories and land surface models over coarse spatial resolution (e.g., > 0.5°) or spatial units (e.g., agro-ecological zones). However, high-resolution CH4 flux estimates capable of capturing the effects of local climate and management practices on emissions, as well as replicating in situ data, remain challenging to produce because of the scarcity of high-resolution maps of paddy-rice and insufficient understanding of CH4 predictors. Here, we combine paddy-rice methane-flux data from 23 global eddy covariance sites and MODIS remote sensing data with machine learning to 1) evaluate data-driven model performance and variable importance for predicting rice CH4 fluxes; and 2) produce gridded up-scaling estimates of rice CH4 emissions at 5000-m resolution across Monsoon Asia, where ∼87% of global rice area is cultivated and ∼ 90% of global rice production occurs. Our random-forest model achieved Nash-Sutcliffe Efficiency values of 0.59 and 0.69 for 8-day CH4 fluxes and site mean CH4 fluxes respectively, with land surface temperature, biomass and water-availability-related indices as the most important predictors. We estimate the average annual (winter fallow season excluded) paddy rice CH4 emissions throughout Monsoon Asia to be 20.6 ± 1.1 Tg yr−1 for 2001–2015, which is at the lower range of previous inventory-based estimates (20–32 CH4 Tg yr−1). Our estimates also suggest that CH4 emissions from paddy rice in this region have been declining from 2007 through 2015 following declines in both paddy-rice growing area and emission rates per unit area, suggesting that CH4 emissions from paddy rice in Monsoon Asia have likely not contributed to the renewed growth of atmospheric CH4 in recent years.

Quantifying global terrestrial photosynthesis is essential to understanding the global carbon cycle and the climate system. Remote sensing has played a pivotal role in advancing our understanding of photosynthesis from leaf to global scale; however, substantial uncertainties still exist. In this review, we provide a historical overview of theory, modeling, and observations of photosynthesis across space and time for decadal intervals beginning in the 1950s. Then we identify the key uncertainties in global photosynthesis estimates, including evaluating light intercepted by canopies, biophysical forcings, the structure of light use efficiency models and their parameters, like photosynthetic capacity, and relationships between sun-induced chlorophyll fluorescence and canopy photosynthesis. Finally, we review new opportunities with big data and recently launched or planned satellite missions.

Continuous monitoring of vegetation indices (VIs) the fraction of absorbed photosynthetically active radiation (fPAR) and leaf area index (LAI) through satellite remote sensing has advanced our understanding of biosphere–atmosphere interactions. Substantial efforts have been put into monitoring individual variables in the field, but options to concurrently monitor VIs, fPAR, and LAI in-situ have been lacking. In this paper, we present the Smart Surface Sensing System (4S), which automatically collects, transfers and processes VIs, fPAR and LAI data streams. The 4S consists of a microcomputer, controller and camera, a multi-spectral spectrometer built in with a light-emitting diode (LED) and an internet connection. Lab testing and field observations in a rice paddy site that experiences wet summer monsoon seasons confirmed the linear response of 4S to light intensities in the blue, green, red and near-infrared spectral channels, with wide ranging temperatures and humidity having only a minor impact on 4S throughout the growing season. Applied over an entire rice growing season (day of year [DOY] 120 - 248), VIs and fPAR from 4S were linearly related to corresponding VIs from a reference spectrometer (R2 = 0.98; NDVI, R2 = 0.96; EVI) and the LAI-2200 instrument (R2 = 0.76), respectively. Integration of gap fraction-based LAI from LED sensors and a green index from the micro-camera allowed tracking of the seasonality of green LAI. The continuous and diverse nature of 4S observations highlights its potential for evaluating satellite remote sensing products. We believe that 4S will be useful for the expansion of ecological sensing networks across multiple spatial and temporal scales.

Sun-induced chlorophyll fluorescence (SiF) is increasingly used as a proxy for vegetation canopy photosynthesis. While ground-based, airborne, and satellite observations have demonstrated a strong linear relationship between SiF and gross primary production (GPP) at seasonal scales, their relationships at high temporal resolution across diurnal to seasonal scales remain unclear. In this study, far-red canopy SiF, GPP, and absorbed photosynthetically active radiation (APAR) were continuously monitored using automated spectral systems and an eddy flux tower over an entire growing season in a rice paddy. At half-hourly resolution, strong linear relationships between SiF and GPP (R2 = 0.76) and APAR and GPP (R2 = 0.76) for the whole growing season were observed. We found that relative humidity, diffuse PAR fraction, and growth stage influenced the relationships between SiF and GPP, and APAR and GPP, and incorporating those factors into multiple regression analysis led to improvements up to R2 = 0.83 and R2 = 0.88, respectively. Relationships between LUEp (=GPP/APAR) and LUEf (=SiF/APAR) were inconsistent at half-hourly and weak at daily resolutions (R2 = 0.24). Both at diurnal and seasonal time scales with half-hourly resolution, we found considerably stronger linear relationships between SiF and APAR than between either SiF and GPP or APAR and GPP. Overall, our results indicate that for subdiurnal temporal resolution, canopy SiF in the rice paddy is above all a very good proxy for APAR at diurnal and seasonal time scales and that therefore SiF-based GPP estimation needs to take into account relevant environmental information to model LUEp. These findings can help develop mechanistic links between canopy SiF and GPP across multiple temporal scales.

Sun-induced chlorophyll fluorescence (SIF) is a promising new tool for remotely estimating photosynthesis. However, the degree to which incoming solar radiation and the structure of the canopy rather than leaf physiology contribute to SIF variations is still not well characterized. Therefore, we investigated relationships between SIF and variables that at least partly capture the canopy structure component of SIF. For this, we relied on high-quality SIF observations from ground-based instruments, high-resolution airborne SIF imagery and the most recent satellite SIF products to cover large ranges in spatial and temporal resolution and diverse ecosystems. We found that the canopy structure-related near-infrared reflectance of vegetation multiplied by incoming sunlight (NIRVP) is a robust proxy for far-red SIF across a wide range of spatial and temporal scales. Our findings indicate that contributions from leaf physiology to SIF variability are small compared to the structure and radiation components. Also, NIRVP captured spatio-temporal patterns of canopy photosynthesis better than SIF, which seems to be mostly due to the greater retrieval noise of SIF. Compared to other relevant structural SIF proxies, NIRVP showed more robust relationships to SIF, especially at the global scale. Our results highlight the promise of using widely available NIRVP data for vegetation monitoring and also indicate the potential of using SIF and NIRVP in combination to extract physiological information from SIF.

International audience; Disentangling the individual contributions from vegetation and soil in measured canopy reflectance is a grand challenge to the remote sensing and ecophysiology communities. Since Solar Induced chlorophyll Fluorescence (SIF) is uniquely emitted from vegetation, it can be used to evaluate how well reflectance-based vegetation indices (VIs) can separate the vegetation and soil components. Due to the residual soil background contributions, Near-infrared (NIR) reflectance of vegetation (NIRv) and Difference Vegetation index (DVI) present offsets when compared to SIF (i.e., the value of NIRv or DVI is non-zero when SIF is zero). In this study, we proposed a simple framework for estimating the true NIR reflectance of vegetation from Hyperspectral measurements (NIRvH) with minimal soil impacts. NIRvH takes advantage of the spectral shape variations in the red-edge region to minimize the soil effects. We evaluated the capability of NIRvH, NIRv and DVI in isolating the true NIR reflectance of vegetation using the data from both the model-based simulations and Hyperspectral Plant imaging spectrometer (HyPlant) measurements. Benchmarked by simultaneously measured SIF, NIRvH has the smallest offset (0–0.037), as compared to an intermediate offset of 0.047–0.062 from NIRv, and the largest offset of 0.089–0.112 from DVI. The magnitude of the offset can vary with different soil reflectance spectra across spatio-temporal scales, which may lead to bias in the downstream NIRv-based photosynthesis estimates. NIRvH and SIF measurements from the same sensor platform avoided complications due to different geometry, footprint and time of observation across sensors when studying the radiative transfer of reflected photons and SIF. In addition, NIRvH was primarily determined by canopy structure rather than chlorophyll content and soil brightness. Our work showcases that NIRvH is promising for retrieving canopy structure parameters such as leaf area index and leaf inclination angle, and for estimating fluorescence yield with current and forthcoming hyperspectral satellite measurements.

Time series of vegetation indices (e.g. normalized difference vegetation index [NDVI]) and color indices (e.g. green chromatic coordinate [GCC]) based on radiometric measurements are now available at different spatial and temporal scales ranging from weekly satellite observations to sub-hourly in situ measurements by means of near-surface remote sensing (e.g. spectral sensors or digital cameras). In situ measurements are essential for providing validation data for satellite-derived vegetation indices. In this study we used a recently developed method to calculate NDVI from near-infrared (NIR) enabled digital cameras (NDVIC) at 17 sites (for a total of 74 year-sites) encompassing six plant functional types (PFT) from the PhenoCam network. The seasonality of NDVIC was comparable to both NDVI measured by ground spectral sensors and by the moderate resolution imaging spectroradiometer (MODIS). We calculated site- and PFT-specific scaling factors to correct NDVIC values and recommend the use of site-specific NDVI from MODIS in order to scale NDVIC. We also compared GCC extracted from red-green-blue images to NDVIC and found PFT-dependent systematic differences in their seasonalities. During senescence, NDVIC lags behind GCC in deciduous broad-leaf forests and grasslands, suggesting that GCC is more sensitive to changes in leaf color and NDVIC is more sensitive to changes in leaf area. In evergreen forests, NDVIC peaks later than GCC in spring, probably tracking the processes of shoot elongation and new needle formation. Both GCC and NDVIC can be used as validation tools for the MODIS Land Cover Dynamics Product (MCD12Q2) for deciduous broad-leaf spring phenology, whereas NDVIC is more comparable than GCC with autumn phenology derived from MODIS. For evergreen forests, we found a poor relationship between MCD12Q2 and camera-derived phenology, highlighting the need for more work to better characterize the seasonality of both canopy structure and leaf biochemistry in those ecosystems. Our results demonstrate that NDVIC is in excellent agreement with NDVI obtained from spectral measurements, and that NDVIC and GCC can complement each other in describing ecosystem phenology. Additionally, NDVIC allows the detection of structural changes in the canopy that cannot be detected by visible-wavelength imagery.

Remote sensing of far-red sun-induced chlorophyll fluorescence (SIF) has emerged as an important tool for studying gross primary productivity (GPP) at the global scale. However, the relationship between SIF and GPP at the canopy scale lacks a clear mechanistic explanation. This is largely due to the poorly characterized role of the relative contributions from canopy structure and leaf physiology to the variability of the top-of-canopy, observed SIF signal. In particular, the effect of the canopy structure beyond light absorption is that only a fraction (fesc) of the SIF emitted from all leaves in the canopy can escape from the canopy due to the strong scattering of near-infrared radiation. We combined rice, wheat and corn canopy-level in-situ datasets to study how the physiological and structural components of SIF individually relate to measures of photosynthesis. At seasonal time scales, we found a considerably strong positive correlation (R2 = 0.4–0.6) of fesc to the seasonal dynamics of the photosynthetic light use efficiency (LUEP), while the estimated physiological SIF yield was almost entirely uncorrelated to LUEP both at seasonal and diurnal time scales, with the partial exception of wheat. Consistent with these findings, the canopy structure and radiation component of SIF, defined as the product of APAR and fesc, explained the relationship of observed SIF to GPP and even outperformed GPP estimation based on observed SIF at two of the three sites investigated. These results held for both half-hourly and daily mean values. In contrast, the total emitted SIF, obtained by normalizing observed SIF for fesc, improved only the relationship to APAR but considerably decreased the correlation to GPP for all three crops. Our findings demonstrate the dominant role of canopy structure in the SIF-GPP relationship and establish a strong, mechanistic link between the near-infrared reflectance of vegetation (NIRV) and the relevant canopy structure information contained in the SIF signal. These insights are expected to be useful in improving remote sensing based GPP estimates.

Solar-induced chlorophyll fluorescence (SIF) provides us with new opportunities to understand the physiological and structural dynamics of vegetation from leaf to global scales. However, the relationships between SIF and gross primary productivity (GPP) are not fully understood, which is mainly due to the challenges of decoupling structural and physiological factors that control the relationships. Here, we report the results of continuous observations of canopy-level SIF, GPP, absorbed photosynthetically active radiation (APAR), and chlorophyll: carotenoid index (CCI) in a temperate evergreen needleleaf forest. To understand the mechanisms underlying the relationship between GPP and SIF, we investigated the relationships of light use efficiency (LUEp), chlorophyll fluorescence yield (ΦF), and the fraction of emitted SIF photons escaping from the canopy (fesc) separately. We found a strongly non-linear relationship between GPP and SIF at diurnal and seasonal time scales (R2 = 0.91 with a hyperbolic regression function, daily). GPP saturated with APAR, while SIF did not. Also, there were differential responses of LUEp and ΦF to air temperature. While LUEp reached saturation at high air temperatures, ΦF did not saturate. We found that the canopy-level chlorophyll: carotenoid index was strongly correlated to canopy-level ΦF (R2 = 0.84) implying that ΦF could be more closely related to pigment pool changes rather than LUEp. In addition, we found that the fesc contributed to a stronger SIF-GPP relationship by partially capturing the response of LUEp to diffuse light. These findings can help refine physiological and structural links between canopy-level SIF and GPP in evergreen needleleaf forest.

Several global gross primary production (GPP) and evapotranspiration (ET) remote sensing products exist, mainly provided by machine-learning (e.g. MPI-BGC) and semi-empirical (e.g. MODIS) approaches. Process-based approaches have the advantage of representing the atmosphere-vegetation-soil system and associated fluxes as an organic integration, but their sophistication results in a lack of high spatiotemporal resolution continuous products. Targeting this gap, we reported a new set of global 8-day composite 1-km resolution GPP and ET products from 2000 to 2015, using a simplified process-based model, the Breathing Earth System Simulator (BESS). BESS couples atmosphere and canopy radiative transfer, photosynthesis and evapotranspiration, and uses MODIS atmosphere and land data and other satellite data sources as inputs. We evaluated BESS products against FLUXNET observations at site scale (total of 113 sites, 742 site years), and against MPI-BGC products at global scale. At site scale, BESS 8-day products agreed with FLUXNET observations with R 2 = 0.67 and RMSE = 2.58 gC m − 2 d − 1 for GPP, and R 2 = 0.62 and RMSE = 0.78 mm d − 1 for ET, respectively, and they captured a majority of seasonal variability, about half of spatial variability, and a minority of interannual variability in FLUXNET observations. At global scale, BESS mean annual sum GPP and ET maps agreed with MPI-BGC products with R 2 = 0.93 and RMSE = 229 gC m − 2 y − 1 for GPP, and R 2 = 0.90 and RMSE = 118 mm y − 1 for ET, respectively. Over the period of 2001–2011, BESS quantified the mean global GPP and ET as 122 ± 25 PgC y − 1 and 65 × 10 3 ± 11 × 10 3 km 3 y − 1 , respectively, with a significant ascending GPP trend by 0.27 PgC y − 2 ( p

Digital cover photography (DCP) has emerged as an indirect method to measure gap fraction of vegetation canopies. However, as with other photographic methods, determining camera relative exposure value (REV) and threshold for pixel classification, cause substantial uncertainties in gap fraction estimates. Here we propose a new method to improve the measurement of gap fraction under various solar zenith angles (SZAs), sky conditions, and canopy structures. This method computes gap fractions of ambiguous vegetation or sky pixels using an unsaturated raw image from DCP and a reconstructed sky image from the raw image, thus taking full advantage of the potential of raw image processing. This is combined with pre-classification of pixels that are unambiguously canopy and sky to greatly reduce light scattering effects that are likely to be present within the canopy. To test the sensitivity of the new method, we acquired images at one-hour intervals between 20 and 85° of SZAs under closed, half-closed, and open canopies with REV settings from 0 to −5. The new method showed little variation in gap fractions across the diverse SZAs in closed, half-closed, and open canopies. A perforated panel experiment, which was used to test the accuracy of the estimated gap fractions, confirmed that the new method accurately estimated gap fractions across a range of hole size, gap fractions and SZAs. We conclude that the new method opens new opportunities to estimate gap fractions accurately regardless of solar positions from open to closed canopies, and is a significant advance for accurate and precise monitoring of canopy cover and leaf area index (LAI), and for calibrating and evaluating satellite remote sensing LAI products.

Satellite image fusion methods that improve spatial and temporal resolution have significant potential to advance understanding of ecosystem dynamics in space and time. However, systematic evaluations of image fusion methods against in situ spectral data are lacking. Here, we used a suite of in situ spectral data collected at 60 elementary sampling units (10 × 10 m) covering 15 Landsat pixel (30 × 30 m) plots and one Moderate Resolution Imaging Spectroradiometer (MODIS) pixel (250 × 250 m) throughout the entire growing season in a heterogeneous rice paddy landscape to evaluate four state-of-the-art image fusion NDVI products. They include the Enhanced Spatial and Temporal Adaptive Reflectance Fusion Model (ESTARFM), Flexible Spatiotemporal DAta Fusion (FSDAF), SaTellite dAta IntegRation (STAIR), and the CubeSat Enabled Spatio-Temporal Enhancement Method (CESTEM); the former three blended Landsat and MODIS data, whereas the latter combined CubeSats, Landsat, and MODIS observations. All fusion products showed strong linear relationships against in situ data when combining all spatial and temporal observations (R2: 0.73 to 0.93) although there were partly negative biases (–1% to –9%). These biases resulted from forcing data to image fusion algorithms, such as Landsat (–4%) and MODIS (–7%). Performance difference between fusion methods were considerably larger for spatial than for temporal variation. Furthermore, Landsat NDVI explained only 17–22% of spatial variation against in situ spectral data, which can be translated into weak performance of image fusion products to predict spatial variability in NDVI. Image fusion products that relied on spatial interpolation showed large biases (–15% to –30%) for a vegetation plot surrounded by mixed land cover plots. Our results highlight key sources of uncertainty and will be instrumental in improving satellite image fusion methods to monitor land surface phenology in space and time.

International audience; Terrestrial gross primary productivity (GPP) plays an essential role in the global carbon cycle, but the quantification of the spatial and temporal variations in photosynthesis is still largely uncertain. Our work aimed to investigate the potential of remote sensing to provide new insights into plant photosynthesis at a fine spatial resolution. This goal was achieved by exploiting high-resolution images acquired with the FLuorescence EXplorer (FLEX) airborne demonstrator HyPlant. The sensor was flown over a mixed forest, and the images collected were elaborated to obtain two independent indicators of plant photosynthesis. First, maps of sun-induced chlorophyll fluorescence (F), a novel indicator of plant photosynthetic activity, were successfully obtained at both the red and far-red peaks (r2 = 0.89 and p 0.05). The spatial relationships found at high resolution provide valuable insight into the critical role of spatial heterogeneity in controlling the relationship between the far-red F and the GPP, indicating the need to consider this heterogeneity at a coarser resolution.

Clumping index quantifies the level of foliage aggregation, relative to a random distribution, and is a key structural parameter of plant canopies and is widely used in ecological and meteorological models. In this study, the inter- and intra-annual variations in clumping index values, derived from the Moderate Resolution Imaging Spectroradiometer (MODIS) BRDF product, are investigated at six forest sites, including conifer forests, a mixed deciduous forest and an oak-savanna system. We find that the clumping index displays large seasonal variation, particularly for the deciduous sites, with the magnitude in clumping index values at each site comparable on an intra-annual basis, and the seasonality of clumping index well captured after noise removal. For broadleaved and mixed forest sites, minimum clumping index values are usually found during the season when leaf area index is at its maximum. The magnitude of MODIS clumping index is validated by ground data collected from 17 sites. Validation shows that the MODIS clumping index can explain 75% of variance in measured values (bias = 0.03 and rmse = 0.08), although with a narrower amplitude in variation. This study suggests that the MODIS BRDF product has the potential to produce good seasonal trajectories of clumping index values, but with an improved estimation of background reflectance.

article Vegetation indices are important remotely sensed metrics for ecosystem monitoring and land surface process assessment, among which Normalized Difference Vegetation Index (NDVI) has been most widely used. The newly launched Landsat 8 Operational Land Imager (OLI) sensor, together with its predecessor Landsat 7 Enhanced Thematic Mapper Plus (ETM+), provides continuous earth observations with an 8-day interval. The design improvements of the new sensor, including narrower near-infrared waveband, higher signal-to-noise ratio (SNR), and greater radiometric sensitivity highlight the need for investigating the land surface observation properties, especially its consistency with data from its predecessors and other satellite sensors. This study aims to evaluate the characteristics of Landsat 8 OLI-derived NDVI against Landsat 7 ETM+ by cross-comparison and by comparing with Moderate Resolution Imaging Spectroradiometer (MODIS) and Geostationary Ocean Color Imager (GOCI)-derived NDVIs as well as in-situ NDVI measurements. Simulations of Top of Atmosphere (TOA) reflectance and surface reflectance of broadleaf trees and water were conducted for Landsat 8 OLI, Landsat 7 ETM+, and MODIS in order to evaluate the impact of band pass difference on NDVI calculation. Four consecutive pairs of Landsat 8 OLI and Landsat 7 ETM+ data over China and Korea were examined, and NDVIs derived from TOA reflectance and surface reflectance by three atmospheric correction methods were evaluated. Both simula- tions and comparisons showed that NDVIs derived from atmospherically-corrected surface reflectance had good consistency, while the simulation showed that the agreement varied with atmospheric characteristics. The four pairs of Landsat 8 OLI and Landsat 7 ETM+ NDVI had a mean bias error within ±0.05, and R 2 from 0.84 to 0.98.

a b s t r a c t Estimates of the canopy gap fraction, on which calculations of leaf area index (LAI) are based, are sensi- tive to photographic exposure in upward-facing images. In this article we describe a simple, automated method of image acquisition and processing that eliminates both subjectivity and the need for the oper- ator to consider photographic exposure in the field. A key strength of our methodology was the use of a test apparatus (perforated aluminum screen) with a precisely known gap fraction; this allowed us to separate the confounding effects of gap size and gap fraction on the optimal photographic exposure for a canopy. We took photographs of the test apparatus at different photographic exposures; we varied the gap fraction by covering a proportion of the holes in the screen, and also varied gap size by varying the distance of the camera from the screen. We acquired both raw images and JPEG images. We found that the optimal exposure of JPEG images varied with both gap size and gap fraction, not just gap fraction as previously assumed. Underexposing by one stop yielded raw data that were never clipped resulting in data loss, but that used most of the 14-bit range of the raw file. We also found that it was easily possible to standardize photographic exposure during image processing by acquiring raw images in the field; thus eliminating the variation in estimated gap fraction and LAI associated with exposure variations. This result was replicated in both fisheye images and cover images that we acquired in real canopies. We recommend the following protocol for acquiring canopy images in future studies

The Moderate Resolution Imaging Spectroradiometer (MODIS) instrument’s land and atmosphere data are important to many scientific analyses that study processes at both local and global scales. The Terra and Aqua MODIS satellites acquire data of the entire Earth’s surface every one or two days in 36 spectral bands. MODIS data provide information to complement many of the ground-based observations but are extremely critical when studying global phenomena such as gross photosynthesis and evapotranspiration. However, data procurement and processing can be challenging and cumbersome due to difficulties in volume, size of data and scale of analyses. For example, the very first step in MODIS data processing is to ensure that all products are in the same resolution and coordinate system. The reprojection step involves a complex inverse gridding algorithm and involves downloading tens of thousands of files for a single year that is often infeasible to perform on a scientist’s desktop. Thus, use of large-scale resource environments such as high performance computing (HPC) environments are becoming crucial for processing of MODIS data. However, HPC environments have traditionally been used for tightly coupled applications and present several challenges for managing data-intensive pipelines. We have developed a data-processing pipeline that downloads the MODIS swath products and reprojects the data to a sinusoidal system on an HPC system. The 10 year archive of the reprojected data generated using the pipeline is made available through a web portal. In this paper, we detail a system architecture (CAMP) to manage the lifecycle of MODIS data that includes procurement, storage, processing and dissemination. Our system architecture was developed in the context of the MODIS reprojection pipeline but is extensible to other analyses of MODIS data. Additionally, our work provides a framework and valuable experiences for future developments and deployments of data-intensive pipelines from other scientific domains on HPC systems.

article i nfo Emerging near-surface remote sensing techniques have advanced our ability to monitor forest canopy phenolo- gy. Thus far, however, little effort has been made to monitor the phenologies of the various canopies of multi- layer forestsseparately,despite theirimportance inregulating forestbiogeochemicalcycles. Here we reportphe- nological changesinmulti-layercanopiesofdeciduousbroadleafandevergreenneedleleaf forestsintheRepublic of Korea during the spring of 2013. We installed light-emitting diode (LED) sensors at four different canopy heights at two sites to measure the normalized difference vegetation index (NDVI) using red and near-infrared (NIR) spectral reflectance and to estimate leaf area index (LAI) using the blue band gap fraction. LED-sensors identified leaf-out dates of over- and understory canopies at both sites; leaves unfolded 8-11 days earlier in the understory canopy than the overstory canopy. At the deciduous forest site, LED-NDVI failed to capture the leaf-out date in the overstory canopy, because all four LED-sensors started to see green-up from the understory canopy while the overstory canopy was leafless. LED-LAI identified different leaf-out dates for the over- and un- derstory canopy, because the gap fraction was measured explicitly for each canopy layer. In the evergreen forest site, LED-NDVI signals between the top of the tower and beneath the overstory canopy were decoupled be- cause of the dense evergreen overstory canopy. Both LED-NDVI and LED-LAI identified new needle expan- sion in the overstory canopy and understory canopy development. MODIS NDVI agreed well with LED-NDVI data (R 2 = 0.96, RMSE = 0.04) at the deciduous forest site, and we discovered that understory canopy de- velopment determined the onset of greenness based on MODIS NDVI data. LED-LAI data agreed well with independent estimates from the other instruments, indicating that LED-sensors could be used to monitor multi-layer canopy LAI. Continuous, in-situ observation of multi-layer canopy phenology will aid in the in- terpretation of satellite remote sensing phenology products and improve land surface models that adopt a multi-layer canopy model.

Statistical gross primary productivity (GPP) estimation from remote sensing observations has mostly been attempted on the basis of multispectral observations. To make full use of the information contained in vegetation spectra, however, hyperspectral observations should be used in combination with appropriate multivariate methods. Nevertheless, only very few previous studies attempted to estimate GPP directly from hyperspectral observations and did so on the basis of reflectance, with only a limited number of temporally discontinuous observations. In this study, we used long-term, continuous, half-hourly hyperspectral observations covering the visible and near-infrared spectral range to estimate GPP directly from upwelling irradiance using partial least square (PLS) regression in a rice paddy. To gain a better understanding of processes underlying the PLS estimation, we used extensive complementary field observations to run process-based simulations using the SCOPE model. We then applied PLS regression to the simulated hyperspectral data in the same way as for the observations and disentangled contributions related to relevant physiological processes, namely sun-induced chlorophyll fluorescence (SIF) and xanthophyll cycle-related spectral changes (XC). We found that upwelling hyperspectral irradiance in the visible and near-infrared spectral range predicted GPP better than reflectance. Furthermore, PLS-based GPP estimates outperformed both far-red SIF and widely used vegetation index-based methods. However, the most relevant information for the observation-based PLS-models was not clearly related to XC or SIF as the near-infrared spectral range showed comparable performance. Also, the simple average of upwelling irradiance over the 850–900 nm range outperformed the other non-multivariate approaches, including far-red SIF. These results held for the evaluation in terms of the seasonal variation of GPP, while there was apparently a small contribution of SIF and XC for the diurnal variation. The simulation-based analysis showed that SIF and XC contributed useful information to both GPP and photosynthetic light use efficiency (LUE) estimates at both seasonal and diurnal time scales. The strongest unique contribution from either SIF or XC, however, was to the diurnal variation of GPP and XC showed considerably better performance than SIF. We did not find improvements when combining the spectral regions of XC (500–570 nm) and SIF (650–800 nm) to estimate GPP. While SIF showed improvements when combined with the remaining spectral information excluding XC, this was not the case for XC. Our approach combines the strengths of process-based modeling with multivariate statistical analysis to improve our understanding of the usable information content in vegetation spectra and is highly relevant for further developing suitable methods for GPP estimation at large scales.

Solar-induced chlorophyll fluorescence (SIF) has emerged as a leading approach for remote sensing of gross primary productivity (GPP). While SIF has an intrinsic, underlying relationship with canopy light capture and light use efficiency, these physiological relationships are obscured by the fact that satellites observe a small and variable fraction of total emitted canopy SIF. Upon emission, most SIF photons are reabsorbed or scattered within the canopy, preventing their observation remotely. The complexities of the radiative transfer process, which vary across time and space, limit our ability to reliably infer physiological processes from SIF observations. Here, we propose an approach for estimating the fraction of total emitted near-infrared SIF (760 nm) photons that escape the canopy by combining the near-infrared reflectance of vegetation (NIRV) and the fraction of absorbed photosynthetically active radiation (fPAR), two widely available remote sensing products. Our approach relies on the fact that NIRV is resilient against soil background contamination, allowing us to reliably calculate the bidirectional reflectance factor of vegetation, which in turn conveys information about the escape ratio of SIF photons. Our NIRV-based approach explains variations in the escape ratio with an R2 of 0.91 and an RMSE of 1.48% across a series of simulations where canopy structure, soil brightness, and sun-sensor-canopy geometry are varied. The approach is applicable to conditions of low leaf area index and fractional vegetation cover. We show that correcting for the escape ratio of SIF using NIRV provides robust estimates of total emitted SIF, providing for the possibility of studying physiological variations of fluorescence yield at the global scale.

Clumping index, the measure of foliage grouping relative to a random distribution of leaves in space, is a key structural parameter of plant canopies that influences canopy radiation regimes and controls canopy photosynthesis and other land–atmosphere interactions. In this study, we retrieve the clumping index using the original 275 m resolution data of the Multi-angle Imaging SpectroRadiometer (MISR) instrument over a set of sites representing diverse biomes and different canopy structures. Also for the first time, the MISR derived clumping index values are directly validated with both in-situ vertical profiles and seasonal trajectories of clumping index. Our results illustrate that MISR data with 275 m allow clumping index estimates at much more pertinent scales (both spatial and temporal) than previous maps from Polarization and Directionality of Earth Reflectances (POLDER) and Moderate Resolution Imaging Spectroradiometer (MODIS) for modeling local carbon and energy fluxes.

Correct estimates of gap fraction are essential for quantifying canopy architectural variables, such as leaf area and clumping indices, which modify land–atmosphere interactions. However, gap fraction measurements from optical sensors are contaminated by radiation that is scattered by plant elements and ground surfaces. In this study, we propose a simple one-dimensional, invertible, bidirectional transmission model to remove scattering effects from gap fraction measurements. To evaluate how well the proposed model computes scattered radiance under a variety of ecosystem conditions, we compared simulated scattered radiance by the proposed model to a more sophisticated three-dimensional model in four ecosystem types (oak-grass savanna, birch, pine, and spruce stands). The simple model showed good agreement with the three-dimensional model in the scattering factor (scattered radiation from leaves normalized by sky radiation), except for highly reflective stems such as birch. The simple model showed that the scattering factor is highest when the leaf area index (LAI) is low (1–2 m 2 m −2 ) in a non-clumped canopy, potential errors in estimating the LAI increase with an increase in LAI, and bright land surfaces (e.g., snow and bright soil) and bright stems (e.g., birch) can contribute significantly to scattering effects. By applying the simple model with LAI-2200 data collected in an oak-grass savanna woodland, we found that the scattering factor causes significant underestimation of the LAI (up to 26% for sunny conditions, 7.7% for diffuse sky conditions) and significant overestimation of the apparent clumping index (up to 14% for sunny conditions, 4.3% for diffuse sky conditions). The LAI is underestimated because of the effect of scattered radiation on gap fraction estimates, which cause overestimation of the clumping index. Even under highly diffuse sky conditions, errors in LAI estimates due to scattering effects are not always negligible (up to 7.7% underestimation). The proposed inversion scheme provides an opportunity to quantify gap fractions, LAI, and apparent clumping index even under sunny conditions.

Most land surface and ecosystem models assume that a vegetated canopy can be abstracted as a turbid medium when such models compute mass, energy, and carbon exchange. However, those models fail to simulate radiation environments in heterogeneous landscapes. This study aims to couple a spatially explicit three-dimensional (3D) shortwave and longwave radiative transfer model with a soil and canopy energy balance and canopy physiology model (CANOAK-FLiES), to investigate how well the 3D model performs in a heterogeneous landscape compared to a 1D model. The canopy structural parameters were extracted using airborne-based Light Detection and Ranging (LiDAR) data and digital cover photographs. The developed model was compared with a wide variety of field and remote sensing data (e.g., hyper-spectral remote sensing images, overstory and understory radiation, and tower-based energy and CO2 fluxes) in a woody oak savanna in CA, USA. Overall, the simulated spatial and diurnal patterns of the radiation environment were consistent with the measurements of shortwave and longwave wavebands. The 3D approach worked better than the 1D approach from the wet mild spring to the dry hot summer while the relative importance of using the 3D approach depends on climate and canopy physiological conditions as well as the canopy structure. With low leaf area index in the oak woodland, the woody elements absorbed 12%–39% of the PAR and 20%–52% of the NIR radiation. Consequently, 12% of the daytime energy flux was used as woody heat storage at our study site. This novel 3D model has the potential to serve as a useful tool for analyzing the spatio-temporal variability of radiation and energy fluxes in evaluating footprints of radiation sensors and eddy covariance fluxes, and serve as a standard in evaluating the performance of a hierarchy of simpler land surface models to compute mass and energy exchange of heterogeneous landscapes.

The continued spread of invasive weeds is threatening ecosystem health throughout North America. Understanding the relationships between invasive weeds’ key phenological phases and structural and/or functional canopy development is an essential step for making informed decisions regarding their management. We analyzed a three-year image archive obtained from an inexpensive webcam overlooking a perennial pepperweed (Lepidium latifoliumL.) infestation in California to explore the ability of red (R)‐green (G)‐blue (B) color space information to track the structural and functional development of the pepperweed.Wecharacterizedstructuralandfunctionalcanopydevelopmentthroughsurfaceroughness length (z 0m ; a proxy for canopy height and leaf area index) and canopy photosynthesis (F A ), respectively, both of which we derived from eddy covariance measurements. Here we demonstrate the use of cross-correlation functions to determine the temporal lags between chromatic coordinates and two color indices, all calculated fromRGBbrightness levels, withz 0m andF A . We found that these color metrics fail to represent the structural and/or functional state of the canopy. In contrast,relative luminance(CIE Y) appears to be a better indicator forz 0m and especially forF A . We calculatedCIE Yfrom pepperweedRGB brightness levels in relation to hypothetical horizontal referenceRGBbrightness levels. We obtained the latter by applying the ratio between horizontally measured and hypothetical incoming solar radiation on a vertical surface toRGBbrightness levels of a vertically oriented reference of invariant light-grey color. We conclude that webcam image archives may provide an inexpensive tool for making informed decisions regarding the timing but not for assessing the effectiveness of invasive plant control measures such as mowing.

The effectiveness of using leveled digital camera for measuring leaf inclination angles was investigated in this study as an inexpensive and convenient alternative to existing approaches. The new method is validated with manual leaf angle measurements for various broadleaf tree species common to hemi-boreal region of Estonia and the tropical forests of Hawai’i Islands. The acquired leaf angle distributions suggest that planophile case might be more appropriate than the commonly assumed spherical as the general approximation of leaf orientation while modeling the radiation transmission through the canopies of (hemi)-boreal broadleaf stands. However, direct leaf inclination measurements should be obtained whenever possible, as there will always exist a large variety of leaf orientation, both among different species and in the space–time domain within a single species. The camera method tested in this study provides a new robust and affordable tool to obtain this information.

We developed and tested the use of light emitting diodes (LEDs) to monitor vegetation reflectance in narrow spectral bands as a tool suitable for quantifying and monitoring ecosystem structure, function and metabolism. LEDs are appealing because they are inexpensive, small and reliable light sources that, when used in reverse mode, can measure spectrally selective radiation. We selected LEDs in red and near-infrared wavebands as they are used to calculate the normalized difference vegetation index (NDVI). The lab experiments revealed that the LEDs showed linear relation with a hyper-spectral spectroradiometer (R2 > 0.94 and 0.99 for red and NIR, respectively) and marginal sensitivity to temperature. To test the efficacy of this novel sensor, we measured spectral reflectance with LEDs over an annual grassland in California over 3.5 years. The LED-sensor captured daily to interannual variation of the spectral reflectance at the two bands with reliable and stable performance. The spectral reflectance in the two bands and NDVI proved to be useful to identify the leaf-on and leaf-off dates (mean bias errors of 5.3 and 4.2 days, respectively) and to estimate canopy photosynthesis (R2 = 0.91). We suggest that this novel instrument can monitor other structural and functional (e.g. leaf area index, leaf nitrogen) variables by employing LEDs that have other specific wavelengths bands. Considering that off-the-shelf LEDs cover a wide range of wavebands from the ultraviolet to near-infrared regions, we believe that the research community could explore a range of similar instruments across a range of bands for a variety of ecological applications.

Many studies on land surface radiation balances have relied on geostationary satellites. These satellites have provided data with high temporal resolution (less than 3 h); however, the spatial resolution was too coarse (20–250 km scale) to investigate local-scale land surface radiation balances. Moderate resolution imaging spectroradiometer (MODIS) – onboard both the Terra and Aqua satellites – yields a tradeoff with regard to this problem by providing higher spatial resolution (∼1 km scale) and sensing all over the earth nearly twice a day during daytime; this provides a potential tool for the periodical monitoring of the land surface energy balance. The reliability of MODIS-derived estimates is, however, affected by the presence of multiple error sources, such as those related to heterogeneous land cover and complex topography. In this study, we have used atmospheric (5 and 10 km scale) and land (1 km scale) products obtained from both the Terra and Aqua MODIS devices as inputs in order to estimate the radiation components (1 km scale) under clear daytime conditions over a heterogeneous farmland area and a rugged deciduous forest in the Korea Flux Network (KoFlux). The reliability of these estimates and the associated errors were evaluated by comparing against field measurements taken for 41 and 26 clear days with regard to the farmland and forest sites, respectively. Solar radiation was successfully retrieved with a root mean square error (RMSE) of ∼20 W m−2 for both the Terra and Aqua devices over the flat farmland site, whereas the rugged forest site exhibited corresponding values of 40 and 65 W m−2 RMSE values with consistent positive biases (presumably caused by topographic effects). The RMSE values of the downward longwave radiation were ∼20 W m−2 for both the Terra and Aqua devices for both these sites. The sensitivities of the upward components of the shortwave and longwave radiations varied with the RMSE values to the scale of the spatial heterogeneity of both the sites. Consequently, the RMSE values of the net radiation ranged from 33 to 61 W m−2 for both the devices at both the sites. Our results suggest that the scales of the patch mosaics within the landscapes need to be quantified for proper retrieval of the MODIS-derived radiation products. More extensive validation efforts are required to identify and account for major error sources across diverse land surface conditions.

International audience; Interannual variability in biosphere-atmosphere exchange of CO2 is driven by a diverse range of biotic and abiotic factors. Replicating this variability thus represents the ‘acid test’ for terrestrial biosphere models. Although such models are commonly used to project responses to both normal and anomalous variability in climate, they are rarely tested explicitly against inter-annual variability in observations. Herein, using standardized data from the North American Carbon Program, we assess the performance of 16 terrestrial biosphere models and 3 remote sensing products against long-term measurements of biosphere-atmosphere CO2 exchange made with eddy-covariance flux towers at 11 forested sites in North America. Instead of focusing on model-data agreement we take a systematic, variability-oriented approach and show that although the models tend to reproduce the mean magnitude of the observed annual flux variability, they fail to reproduce the timing. Large biases in modeled annual means are evident for all models. Observed interannual variability is found to commonly be on the order of magnitude of the mean fluxes. None of the models consistently reproduce observed interannual variability within measurement uncertainty. Underrepresentation of variability in spring phenology, soil thaw and snowpack melting, and difficulties in reproducing the lagged response to extreme climatic events are identified as systematic errors, common to all models included in this study.

a b s t r a c t The regular monitoring of evapotranspiration from satellites has been limited because of discontinu- ous temporal coverage, resulting in snapshots at a particular point in space and time. We developed a temporal upscaling scheme using satellite-derived instantaneous estimates of evapotranspiration to produce a daily-sum evapotranspiration averaged over an 8-day interval. We tested this scheme against measured evapotranspiration data from 34 eddy covariance flux towers covering seven plant functional types from boreal to tropical climatic zones. We found that the ratio of a half-hourly-sum of potential solar radiation (extraterrestrial solar irradiance on a plane parallel to the Earth's surface) between 10:00 hh and 14:00 hh to a daily-sum of potential solar radiation provides a robust scaling factor to convert a half-hourly measured evapotranspiration to an estimate of a daily-sum; the estimated and measured daily sum evapotranspiration showed strong linear relation (r 2 = 0.92) and small bias (−2.7%). By com- parison, assuming a constant evaporative fraction (the ratio of evapotranspiration to available energy) during the daytime, although commonly used for temporal upscaling, caused 13% underestimation of evapotranspiration on an annual scale. The proposed temporal upscaling scheme requires only latitude, longitude and time as input. Thus it will be useful for developing continuous evapotranspiration estimates in space and time, which will improve continuous monitoring of hydrological cycle from local to global scales.