Peak Of Growing Season Research Articles

A new transferable deep learning approach for crop mapping

ABSTRACT Accurate crop mapping provides important information for government decision-making and agricultural management. Recent advancements in deep learning have significantly enhanced the capabilities of remote sensing-based crop mapping. In this study, we developed a new deep learning approach, DSH, which comprises three modules, DeepLabV3+ (D), channel self-attention (S), and histogram matching (H). The DeepLabV3+ module learns the spatial distribution of crops in the spatial and spectral dimensions. The channel self-attention module was used to enhance the weights of the important features. The histogram-matching module addresses the domain gap between the training and testing images and enhances the transferability of the DSH. In addition, the input data of the DSH are monthly synthesized Sentinel-2 data, thus relaxing the data collection requirements. The performance of the DSH was evaluated and benchmarked against HRNetV2, DeepLabV3+, and random forest (RF) at eight sites in the U.S. China, and France with different farming structures, climate conditions, and landscape complexities. Temporal transfer revealed that the classification performance of DSH improved as the growth season progressed, peaked in the peak growth season (August), and then declined at each study site. The DSH model trained during the peak growth season demonstrated superior temporal transferability. Spatial transfer experiments demonstrated that transferring DSH to sites with similar planting structures achieved higher precision than full spatial transfers. Spatiotemporal transfer experiments conducted at eight sites over three years (2020–2022) demonstrated an average overall accuracy (OA) of 88.1% for DSH, highlighting the importance of similar planting structures for effective spatiotemporal transfer. DSH outperformed HRNetV2, DeepLabV3+, and RF in OA, producer’s accuracy, and user’s accuracy at all eight study sites. Notably, DSH exhibited comparable performance during both the peak and full growth seasons, with an average OA of 93.9%. The classification accuracy of test images after histogram matching remained stable from 2020 to 2022, whereas the accuracy of raw test images fluctuated annually. This indicates that histogram matching is effective in improving the transferability of DSH. Embedding the channel self-attention module into the position of the high-level features had a more significant impact on improving DSH’s performance than other configurations. Visualizing features from the channel self-attention and high-level layers revealed that integrating the channel self-attention module with DeepLabV3+‘s high-level features significantly improved crop mapping by enhancing relevant semantic information. Overall, DSH exhibits robust spatiotemporal generalizability and requires minimal input data, making it suitable for large-scale crop mapping.

Drought shortens subtropical understory growing season by advancing leaf senescence.

Subtropical forests, recognized for their intricate vertical canopy stratification, exhibit high resistance to extreme drought. However, the response of leaf phenology to drought in the species-rich understory remains poorly understood. In this study, we constructed a digital camera system, amassing over 360,000 images through a 70% throughfall exclusion experiment, to explore the drought response of understory leaf phenology. The results revealed a significant advancement in understory leaf senescence phenology under drought, with 11.75 and 15.76 days for the start and end of the leaf-falling event, respectively. Pre-season temperature primarily regulated leaf development phenology, whereas soil water dominated the variability in leaf senescence phenology. Under drought conditions, temperature sensitivities for the end of leaf emergence decreased from -13.72 to -11.06 days °C-1, with insignificance observed for the start of leaf emergence. Consequently, drought treatment shortened both the length of the growing season (15.69 days) and the peak growth season (9.80 days) for understory plants. Moreover, this study identified diverse responses among intraspecies and interspecies to drought, particularly during the leaf development phase. These findings underscore the pivotal role of water availability in shaping understory phenology patterns, especially in subtropical forests.

Seasonal variation of water utilization sources and responses to precipitation in the dominant species of broad-leaved and needle-leaved mixed forests in Mount Lushan, China

Elucidating the seasonal patterns of water sources for dominant species in the sub-tropical humid mountainous forest, analyzing the eco-hydrological complementarity and competition mechanisms among coexisting species, investigating the responses of plant water utilization to precipitation, could provide a theoretical basis for vegetation restoration and management. Based on the stable hydrogen and oxygen isotope technique, we analyzed the δ2H and δ18O characteristics of precipitation, xylem water from Pinus massoniana and Quercus variabilis, and soil water from 0-100 cm depth in Mount Lushan, China. The MixSIAR model, Levins index, and PS index were used to calculate the relative contribution rate of each water source, the hydrological niche breadth, and niche overlap of P. massoniana and Q. variabilis. The results showed that, in the wet season (March to July), P. massoniana primarily utilized soil water from the 0-20 cm and 20-40 cm depths, while Q. variabilis primarily utilized that from the 20-40 cm and 40-60 cm depths. During the dry season (August to September), P. massoniana and Q. variabilis utilized 40-60 cm and 60-80 cm of soil water, respectively, resulting in an increase in the depth of water absorption. In the early growing season (March to April) and the late growing season (September), there was a high hydrological niche overlap between P. massoniana and Q. variabilis, resulting in intensitive water competition. In the middle of the growing season (May to August), the water source was adequately allocated, and the hydrolo-gical niche was segregated to meet the high transpiration demand. Q. variabilis primarily utilized soil water from a depth of 60-80 cm and 60-80 cm before a precipitation event, and from a depth of 0-20 cm and 20-40 cm after the event. In contrast, P. massoniana primarily utilized soil water from a depth of 0-20 cm and 20-40 cm both before and after a precipitation event. In conclusion, water utilization patterns of P. massoniana and Q. variabilis exhibited a seasonal trend, with shallow water uptake during the rainy season and deep water uptake during the dry season. These species are capable of efficiently allocating water resources during the peak growth season, and their root systems actively respond to change in soil moisture level. They have strong adaptability to extreme precipitation events and exhibit remarkable water conservation capabilities.

Multispectral remote sensing approach of predicting the potential distribution and evaluating the current spread of water hyacinth (Eichhornia crassipes)

The water hyacinth is categorized among the world’s top ten worst invasive plant species of aquatic ecosystems. This study assessed changes in the spatiotemporal distributions of the water hyacinth in Lake Koka and Ziway of the Upper Awash River basin during the peak growth season of the plant. Household questionnaires and key informant interviews along with Landsat images for 2013, 2017, and 2021 were collected to identify the past, present, and future potential distributions of the invasive plant in the two lakes. Household surveys and key informant interviews were prepared using the Kobo Toolbox which monitors data collection online. A total number of 413 households were sampled and the data were analyzed through descriptive statistics. For Landsat images, a supervised classification technique was applied to classify the land use classes using the maximum likelihood algorithm. The survey results showed increased water hyacinth expansion in the area since the year 2011. The water hyacinth expansion affected 285 households’ livelihoods by invading 69.0% of their farmlands which caused 97.6% food scarcity in the study districts. The results of the Landsat image indicated that the water hyacinth invasion in Lake Koka occupied 1.48% in 2013, and this increased to 7.13% in 2021, while the water body decreased from 75.94 to 69.90%, respectively. However, the area of the other vegetation is nearly identical between the years 2013–2021. Likewise, water hyacinth covered 4.66% of Lake Ziway in 2013, and this was raised to 8.42% in 2021. At Lake Ziway water hyacinth invasion affected the area of other vegetation coverage as it decreased from 16.19 to 10.67% but the area of the water body remained almost the same during the years. Between 2013 and 2025, the amount of water hyacinths increased in both Lake Koka and Lake Ziway. According to LULC data, the water hyacinth's rate of spread was 0.56% in Lake Koka and 0.95% in Lake Ziway. The results of this study revealed the signals of LULC change due to water hyacinth invasion in the Upper Awash Basin which is considered an important aspect for future water resources planning and management.

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.

Remote Sensing Soil Freeze‐Thaw Status and North American N2O Emissions From a Regional Inversion

AbstractNorth American nitrous oxide (N2O) emissions over 2011–2018 are estimated using the CarbonTracker‐Lagrange regional inversion framework. Emissions are strongest in the Midwestern corn/soybean belt and display a distinct dual maxima seasonal pattern. The first maximum occurs in late winter/early spring, suggestive of freeze‐thaw (FT) effects on denitrification rates and associated N2O emissions. The second maximum occurs in late spring/early summer, consistent with a growing season nitrogen fertilizer‐driven source, although fertilizer applied in late fall may contribute to the FT pulse as well. Interannual variability in the first maximum correlates significantly to soil freeze thaw status derived from remote sensing data. A requisite frozen period in the preceding early winter appears necessary to create conditions for the N2O pulse after thawing. The FT pulse is a prominent feature of the annual cycle in Canadian cropland, where it may be of comparable magnitude to growing season emissions. In contrast, the growing season peak in N2O dominates the FT peak in the Midwestern Corn‐Soybean region of the United States.

Effects of grazing and nitrogen application on greenhouse gas emissions in alpine meadow

Overgrazing and injudicious nitrogen applications have increased emissions of greenhouse gases from grassland ecosystems. To explore the effects and potential mechanisms of grazing, nitrogen application, and their interaction with greenhouse gas (GHG) emissions, field experiments were conducted on the Qinghai-Tibet Plateau for three consecutive years. Alpine meadow plots were subjected to light (8 sheep ha−1) and heavy (16 sheep ha−1) stocking rates, with or without ammonium nitrate (NH4NO3) (90 kg N ha−1 yr−1) treatment to simulate soil nitrogen deposition. During early warm growth season (May–June), peak growth season (July–September), and early cold season (October–November), static-chamber gas chromatography was used to analyze the soil's greenhouse gas emissions (CO2, N2O, and CH4). Results indicated that light stocking rate (LG) led to an increase in cumulative CO2 and N2O emissions, while also promoting CH4 uptake. Conversely, heavy stocking rate (HG) produced contrasting outcomes. Additionally, nitrogen applications significantly increased the short-term CO2 and N2O fluxes peaks. Combined treatment of nitrogen application and light stocking rate (LG + N) resulted in increased CO2 and N2O emissions while decreased CH4 uptake, consequently leading to a significant increase in global warming potential. According to the structural equation model, we discovered that nitrogen application and grazing affected GHG fluxes both directly and indirectly through their impact on the environmental factors. Our findings suggest that in the context of increasing nitrogen deposition in the Qinghai-Tibet Plateau, a moderate increase in stocking rate is more effective than reducing grazing intensity for mitigating global warming potential in alpine meadow.

Spatial-temporal dynamics of land surface phenology over Africa for the period of 1982–2015

Knowledge of the dynamics of vegetation phenology is essential for the understanding of vegetation-climate interactions. Although the interest in phenology study is growing, vegetation phenology in Africa received far less attention compared to the Northern Hemisphere. Africa straddles the northern and southern hemispheres, and the climate has a clear latitudinal gradient, which facilitates the study of the interaction between phenology and climate. In this study, the latitudinal and longitudinal gradients and temporal trends of start of growing season (SOS), peak of growing season (POS), and end of growing season (EOS) were examined using long-term satellite dataset during 1982–2015. The latitudinal variations in these phenology metrics were larger in the northern than those in the southern Africa, especially from 6°N northwards to 16°N. The latitudinal variations in southern Africa had no clear patterns due to the more complex climate systems. For the longitudinal variation, the temporal trends in POS and EOS exhibited a gradient-decreasing rate in northern Africa. Over the period from 1982 to 2015, the overall trends of the phenology in Africa were ‘later SOS’, ‘later POS’, and ‘later EOS’. The faster rate of delay in EOS than in SOS resulted in a prolonged length of growing season (LOS) with 0.50 days/year on average in northern Africa, while a slower rate of delay in EOS than in SOS resulted in a shorter LOS with −0.12 days/year in southern Africa. The prolonged LOS in northern Africa contributes to the increase in the yearly-averaged Normalized Difference Vegetation Index (NDVI) from 1982 to 2000. Nevertheless, the NDVI appeared to have reached saturation around the 2000s, although the LOS was still extending after 2000s. Overall, the findings of this study provide an overall view of the spatial and temporal patterns of land surface phenology in the African continent, and a necessary component for future studies on the response of phenology to climate.

Spring phenology rather than climate dominates the trends in peak of growing season in the Northern Hemisphere.

Shifts in plant phenology regulate ecosystem structure and function, which feeds back to the climate system. However, drivers for the peak of growing season (POS) in seasonal dynamics of terrestrial ecosystems remain unclear. Here, spatial-temporal patterns of POS dynamics were analyzed by solar-induced chlorophyll fluorescence (SIF) and vegetation index in the Northern Hemisphere over the past two decades from 2001 to 2020. Overall, a slow advanced POS was observed in the Northern Hemisphere, while a delayed POS distributed mainly in northeastern North America. Trends of POS were driven by the start of growing season (SOS) rather than pre-POS climate both at hemisphere and biome scale. The effect of SOS on the trends in POS was the strongest in shrublands while the weakest in evergreen broad-leaved forest. These findings highlight the crucial role of biological rhythms rather than climatic factors in exploring seasonal carbon dynamics and global carbon balance.

Impacts of global change on peak vegetation growth and its timing in terrestrial ecosystems of the continental US

Vegetation phenology and peak growth are important indicators reflecting the response of ecosystems to global changes, yet the trends in peak vegetation growth and its drivers have not been well analyzed across different ecosystems. Here, we used time series of satellite-derived Normalized Difference Vegetation Index (NDVI) (1985–2015) to track recent changes in maximum vegetation growth (the maximum of NDVI (NDVImax) and the peak of growing season (POS)) in the continental US, and explored their responses to both climatic drivers (i.e. temperature, precipitation and cloud cover) and atmospheric changes (i.e. nitrogen (N) deposition and atmospheric CO2 concentration [CO2]). We found that precipitation had the greatest impact on NDVImax, accounting for 31.7% of vegetated area, followed by cloud cover (17.1%), temperature (16.4%), [CO2] (13.7%), and N deposition (10.4%), respectively. However, temperature showed the largest contribution to POS change in about 41.9% of vegetated areas, followed by precipitation (18.6%), cloud cover (15.0%), N deposition (6.7%) and [CO2] (5.9%), respectively. Interestingly, the effects of cloud cover highly depended on moisture regime, that is, more cloud cover was associated with a delayed POS and a decreased NDVImax in wetter regions, but could advance POS and increase NDVImax in dry areas. In addition, we also found that an earlier POS was accompanied by a higher NDVImax in most ecosystems, but with lower NDVImax in arid ecosystems. These results reveal the control of both climatic and atmospheric drivers on peak growth, and consequently are of great significance for understanding of the responses and feedbacks of vegetation to global change and improve models of terrestrial ecosystem carbon cycle.

Spatial patterns and driving factors of aboveground and belowground biomass over the eastern Eurasian steppe

The Eurasian steppe is the largest steppe region in the world and is an important part of the global grassland ecosystem. The eastern Eurasian steppe has favorable hydrothermal conditions and has the highest productivity and the richest biodiversity. Located in the arid and semi-arid region, the eastern Eurasian steppe has experienced large-scale grassland degradation due to dramatic climate change and intensive human activities during the past 20 years. Hence, accurate estimation of aboveground biomass (AGB, gC m−2) and belowground biomass (BGB, gC m−2) is necessary. In this study, plenty of AGB and BGB in-situ measurements were collected among dominated grassland types during summer in 2013 and 2016–2018 in the eastern Eurasian steppe. Vegetation indices from the Moderate Resolution Imaging Spectroradiometer (MODIS), Digital Elevation Model (DEM) and climate variables were chosen as independent variables to establish predictive models for AGB and BGB with random forest (RF). Both AGB (R2 = 0.47, MAE = 21.06 gC m−2, and RMSE = 27.52 gC m−2) and BGB (R2 = 0.44, MAE = 173.02 gC m−2, and RMSE = 244.20 gC m−2) models showed acceptable accuracy. Then the RF models were applied to generate spatially explicit AGB and BGB estimates for the study area over the last two decades (2000–2018). Both AGB and BGB showed higher values in the Greater Khingan Mountains and decreased gradually to the east and west sides. The mean values for AGB and BGB were 62.16 gC m−2 and 531.35 gC m−2, respectively. The climatic factors were much more important in controlling biomass than anthropogenic drivers, and shortage of water and raising temperature were the main limiting factor of AGB and BGB, respectively, in the peak growth season. These findings provide scientific data for the scientific management of animal husbandry and can contribute to the sustainable development of grassland ecology in the eastern Eurasian steppe.

Recent trends in the timing of the growing season in New Zealand’s natural and semi-natural grasslands

ABSTRACT We investigate the temporal dynamics of shifts in phenological responses of a range of key stages of the growing season in New Zealand’s three indigenous grassland types over the last 16 years (2001–2016). A near-daily Normalized Difference Vegetation Index (NDVI) time series from MODerate Resolution Imaging Spectroradiometer (MODIS) was used to extract five annual growth phenology indices, namely the Start, End, Length, Peak and Peak NDVI of a growing season. The start of the growing season advanced (i.e. happened earlier) by a median of 7.2, 6.0 and 8.8 days per decade in Alpine, Tall Tussock and Low Producing grassland, whereas the end of the season advanced by a median of 4.5, 0.4 and 0.4 days in the three types respectively. The length of growing season was extended by 3.2, 5.2 and 7.1 days per decade in these three grassland types. Over 86% of the investigated grassland areas showed an advancing (earlier) start of the growing season, and 74% of Alpine grassland showed a trend toward an earlier end of season. Over 63% of all grassland types showed an increase in growing season length. A trend toward earlier growing season peak and overall increasing NDVI in the three grassland types indicate a tendency for increasing vegetation vitality in grassland ecosystems in recent years. The start of growing season was correlated with atmospheric pressure (negatively) and precipitation (positively) changes in winter–spring months, while the timing of the season end is positively correlated with air temperature and solar radiation in summer–autumn months. Our study shows that different grassland types differ in magnitude – but not in direction – of their recent shifts in timing of key growing season stages with high-alpine grasslands showing the strongest response. This study highlights the usefulness of remote sensing for monitoring ecosystem-level phenological shifts over large areas and long time periods.

Ecological performance differs between range centre and trailing edge populations of a cold-water kelp: implications for estimating net primary productivity

Kelp forests are extensive, widely distributed and highly productive. However, despite their importance, reliable estimates of net primary productivity (NPP) are currently unknown for most species and regions. In particular, how performance and subsequent NPP change throughout a species range is lacking. Here, we attempted to resolve this by examining growth and performance of the boreal kelp, Laminaria digitata, from range centre and trailing edge regions in the United Kingdom. During the peak growth season (March/April), range-centre individuals were up to three times heavier and accumulated biomass twice as fast as their trailing-edge counterparts. This was not apparent during the reduced growth season (August/September), when populations within both regions had similar biomass profiles. In total, annual NPP estimates were considerably lower for trailing-edge (181 ± 34 g C m−2 year−1) compared to range-centre (344 ± 33 g C m−2 year−1) populations. Our first-order UK estimates of total standing stock and NPP for L. digitata suggest this species makes a significant contribution to coastal carbon cycling. Further work determining the ultimate fate of this organic matter is needed to understand the overall contribution of kelp populations to regional and global carbon cycles. Nevertheless, we highlight the need for large-scale sampling across multiple populations and latitudes to accurately evaluate kelp species’ contributions to coastal carbon cycling.

Multi-Climate Factors and the Preceding Growth Stage of Vegetation Co-Regulated the Variation of the End of Growing Season in Northeast Inner Mongolia, China

The end of growing season (EOS) is an effective indicator of annual vegetation growth. Previous studies have revealed the dynamics of the EOS with climate change, while the influence of vegetation growth in preceding stage and peak of growing season (POS) on the EOS has not been thoroughly documented. In this study, we used four smoothing methods to obtain EOS dates from the Normalized Difference Vegetation Index (NDVI) in northeast Inner Mongolia (NIM) between 2001–2017, assessed the differences in the spatiotemporal variations of the EOS obtained by the four smoothing methods, and then investigated the impacts of climate factors, summer/ autumn vegetation growth and POS on the EOS. The results showed that the EOS dates obtained with different smoothing methods were broadly consistent in terms of their spatial patterns and temporal trends. In terms of climate factors, the EOS was driven mainly by preseason precipitation for the majority of vegetation types and advanced with increasing precipitation. For the steppe, both minimum temperature (Tmin) and relative humidity (RHU) played the most important roles in regulating the variation of EOS which was delayed with an increase in Tmin and reduction in RHU. Furthermore, our study found an earlier POS and vigorous vegetation growth in summer would jointly advance the steppe EOS, but these relationships were the opposite of each other in meadow and forest regions. Interestingly, the EOS of NIM was more related with vegetation growth in the most recent period before the EOS. This study highlights the importance of ecological processes in the preceding growth stage for understanding the dynamics of EOS.

Sight for Sorghums: Comparisons of Satellite- and Ground-Based Sorghum Yield Estimates in Mali

The advent of multiple satellite systems capable of resolving smallholder agricultural plots raises possibilities for significant advances in measuring and understanding agricultural productivity in smallholder systems. However, since only imperfect yield data are typically available for model training and validation, assessing the accuracy of satellite-based estimates remains a central challenge. Leveraging a survey experiment in Mali, this study uses plot-level sorghum yield estimates, based on farmer reporting and crop cutting, to construct and evaluate estimates from three satellite-based sensors. Consistent with prior work, the analysis indicates low correlation between the ground-based yield measures (r = 0.33). Satellite greenness, as measured by the growing season peak value of the green chlorophyll vegetation index from Sentinel-2, correlates much more strongly with crop cut (r = 0.48) than with self-reported (r = 0.22) yields. Given the inevitable limitations of ground-based measures, the paper reports the results from the regressions of self-reported, crop cut, and (crop cut-calibrated) satellite sorghum yields. The regression covariates explain more than twice as much variation in calibrated satellite yields (R2 = 0.25) compared to self-reported or crop cut yields, suggesting that a satellite-based approach anchored in crop cuts can be used to track sorghum yields as well or perhaps better than traditional measures. Finally, the paper gauges the sensitivity of yield predictions to the use of Sentinel-2 versus higher-resolution imagery from Planetscope and DigitalGlobe. All three sensors exhibit similar performance, suggesting little gains from finer resolutions in this system.

Estimating the peak of growing season (POS) of China’s terrestrial ecosystems

Vegetation phenology is a sensitive indicator of climate change and has substantial impacts on ecosystem carbon uptake. Previous studies have revealed the importance of the start and end of growing season and investigated their dynamics with climate change. However, the interannual variability of the peak of growing season (POS), which also plays a crucial role in carbon cycle of terrestrial ecosystems, has not been well documented. Here, we used field observations from regional flux measurements to evaluate four methods in POS extraction based on time series of satellite observed normalized difference vegetation index (NDVI). Results showed that the combination of Savitzky-Golay filter and cubic spline (SG-cubic spline) method had overall higher accuracy (R2 = 0.77, RMSE = 8.3 days) than current approaches for POS detection. We then analyzed the spatial and temporal patterns of POS in China during 1982–2015 and found an overall advanced trend of POS in most regions, but variations were also observed among biomes. Preseason temperature was the most important factor controlling POS for about 25.6% of all pixels, followed by radiation (19.7%) and precipitation (14.6%), respectively. A higher temperature overall advanced POS, while influences of precipitation and radiation were highly biome-dependent. We further investigated the impacts of POS on plant productivity and found that an advanced POS would increase annual GPP, but this relationship was opposite in arid regions, highlighting the role of soil moisture in dry areas. Our study is meaningful for deepening the understanding between phenology and carbon cycling with respect to future climate change.

Robust Landsat-based crop time series modelling

Reliable satellite monitoring of agriculture is often difficult because surface variations occur rapidly compared to the cloud-free satellite observation frequency. Harmonic time series models, i.e., superimposed sequences of sines and cosines, have an established provenance for fitting satellite vegetation index time series to coarse resolution satellite data, but their application to medium resolution Landsat data for crop monitoring has been limited. Non-linear harmonic models have been shown to perform well over agricultural sites using single-year Moderate Resolution Imaging Spectroradiometer (MODIS) time series, but have not been explored with Landsat data. The 2017 availability of Landsat Analysis Ready Data (ARD) over the United States provides the opportunity to investigate the utility of temporally rich Landsat data for 30 m pixel-level crop monitoring. In this paper, the capability of 5- and 7-parameter linear harmonic models and a 5-parameter non-linear harmonic model applied to a year of Landsat 5 Thematic Mapper (TM) and Landsat 7 Enhanced Thematic Mapper Plus (ETM+) ARD is investigated. The analysis is undertaken over six sites, each defined by a 5000 × 5000 30 m pixel ARD tile, that together include the major conterminous United States (CONUS) crops identified by inspection of the United States Department of Agriculture (USDA) Cropland Data Layer (CDL). The model fits are evaluated as the root mean square difference (RMSD) between the fitted and the observed Landsat data. Considering locations with at least 21 annual Landsat observations, the 7-parameter linear harmonic model (tile mean crop NDVI RMSD values ranging from 0.052 to 0.072) and the 5-parameter non-linear harmonic model (tile mean crop NDVI RMSD values ranging from 0.054 to 0.074) are shown to be able to fit annual Landsat NDVI time series for most CONUS crops, whereas the 5-parameter linear harmonic model cannot (tile mean crop NDVI RMSD values ranging from 0.072 to 0.099). If there are between 15 and 20 annual Landsat observations, the 5-parameter non-linear harmonic model is recommended for fitting annual NDVI crop time series, and if there are ≥21 observations, then either the 5-parameter non-linear or the 7-parameter linear model can be used. The 7-parameter model had marginally smaller mean NDVI RMSD values but larger standard deviations than the 5-parameter non-linear model, likely due to the relative robustness of the non-linear model to over-fitting and oscillations. None of the models could reliably fit crops with multiple stages, such as alfalfa, that are insufficiently sampled using combined Landsat 5 TM and Landsat 7 ETM+ time series. Given the utility of the growing season peak NDVI for crop yield applications, the date and magnitude of the model fitted peak NDVI are compared to quantify model reporting differences. The differences between the 7-parameter linear and the 5-parameter non-linear harmonic models are not large. For each ARD tile, the mean absolute differences in the estimated peak NDVI days varied from <2 days in the northern ARD tiles, which had short growing seasons and similar crops, to less than a week for the other tiles except for nearly 10 days for the California tile that had longer growing seasons and more diverse crops including crops with multiple stages. The paper concludes with a discussion and recommendations for future research.

Tibetan sedges sequester more carbon belowground than grasses: a 13C labeling study

The Kobresia meadows widely distributed on the Tibetan Plateau are very tolerant to intensive grazing. The plant communities in these meadows are currently undergoing a shift in dominance from Cyperaceae (sedges) to Gramineae (grasses) species. We investigated the effects of the replacement of sedges by grasses in an alpine meadow on the allocation of photosynthetic carbon (C) in various plant and soil pools. We selected patches (50 cm × 50 cm) of two vegetation types dominated by sedges (Kobresia capillifolia, K. humilis and Carex spp., accounting for 88% of the total shoot biomass) and grasses (Elymus nutans, Poa crymophila and Leymus secalinus, accounting for 86% of the total shoot biomass), respectively, in an alpine pasture grazed by yaks (Bos grunniens) and sheep (Ovis aries) in the winter season (November–June). Photoassimilate partitioning in various plant and soil pools was compared between sedge- and grass-dominated patches in the peak growth season (August) after labeling of plants in a 13CO2 atmosphere. The sedge-dominated patches had 85% smaller aboveground biomass but a much larger ratio of root (in the 0–50 cm soil depth) to shoot biomass (23.0 under sedges vs. 9.5 under grasses) than the grass-dominated patches. However, similar 13C amounts were assimilated by sedges (1651 mg C m−2) and grasses (1575 mg C m−2) during 2 h of labeling, showing a greater capacity to assimilate C per unit of aboveground biomass in sedges. Over the period of 32 days after labeling, 13C allocation to roots and soil was consistently greater under sedges than under grasses. By day 32 after labeling, the total 13C amount remaining in roots and soil down to 50 cm depth was greater under sedges (78.5% of the total 13C assimilated) than under grasses (65.7%). Although similar C amounts were assimilated by two plant functional groups, sedges transferred more C to roots and soil (through rhizodeposition) than the grasses. This reflects the greater root-to-shoot ratio of sedges than grasses. Consequently, the ongoing replacement of sedges by grasses in plant communities will decrease soil C sequestration in the Kobresia meadows of the Tibetan Plateau.

Water use efficiency and evapotranspiration partitioning for three typical ecosystems in the Heihe River Basin, northwestern China

It is crucial to improve water use efficiency (WUE) and the transpiration fraction of evapotranspiration (T/ET) for water conservation in arid regions. As a link between carbon and water cycling, WUE is defined as the ratio of gross primary productivity (GPP) and ET at the ecosystem scale. By incorporating the effect of vapor pressure deficit (VPD) on WUE, two underlying WUE (uWUE) formulations, i.e. a potential uWUE (uWUEp = GPP·VPD0.5/T) and an apparent uWUE (uWUEa = GPP·VPD0.5/ET), were proposed recently. Both uWUEp and uWUEa can be estimated from eddy covariance measurements, and the ratio of uWUEa and uWUEp was then used to estimate T/ET. This new method for ET partitioning was applied to three typical ecosystems in the Heihe River Basin. Growing season T/ET at the Daman site (0.63) was higher than that at the Arou and Huyanglin sites (0.55) due to the application of plastic film mulching. The effect of leaf area index (LAI) on seasonal variations in T/ET was strong for Arou (R2 = 0.74) and Daman (R2 = 0.76) sites, but weak for Huyanglin (R2 = 0.44) site. Daily T/ET derived using the uWUE method agreed with that using the isotope and lysimeter/eddy covariance methods during the peak growth season at the Daman site. The estimated T using the uWUE method showed consistent seasonal and diurnal patterns and magnitudes with that using the sap flow method at the Huyanglin site. In addition, the uWUE method can estimate ecosystem T/ET without scaling issues, and can effectively capture T/ET variations in relation to LAI changes and the abrupt T/ET changes in response to individual irrigation events. These advantages make the uWUE method effective for ET partitioning at the ecosystem scale, and can be used for water resources management by predicting seasonal pattern of actual plant water requirements to support irrigation strategies in arid regions.

Forest water use in the initial stages of reclamation in the Athabasca Oil Sands Region

AbstractFollowing large‐scale surface oil sands mining, large tracts of the boreal forest in the Athabasca Oil Sands Region of Western Canada are legally required to be reclaimed. A greater understanding of how these novel ecosystems function and develop with regard to water use is crucial to aid in the development of regulatory practices and protocols based on information from ecosystem recovery. In this paper, a 12‐year (2003–2014) eddy covariance measurement record of latent and sensible heat fluxes and gross ecosystem productivity of carbon dioxide is analysed to evaluate how a reclaimed boreal forest has developed during its initial growth period. The study site is a reclaimed oil sands saline‐sodic clay shale overburden deposit that was topped with 100 cm of glacial till and 20 cm of peat mineral mix. The site was seeded with barley (Hordeum spp.) in 2001 to reduce erosion of the soil cover whereas aspen (Populus tremuloides Michx.) and spruce (Picea glauca [Moench] Voss) boreal tree species were planted in 2004. Changes in structure and function corresponded to the transition of dominant vegetation cover from early successional species to forest. Leaf area index increased from a growing season peak of 0.9 in 2003 to 4.0 in 2014 and was associated with an increased growing season gross ecosystem productivity (4.9 to 8.9 g C m−2 day−1), an increased evapotranspiration (1.6 to 3.4 mm day−1), and a decreased partitioning of energy to sensible heat (Bowen's ratio decreased from 1.1 to 0.4). Although canopy conductance increased throughout the 12 years, the shift from early successional species to trees with more conservative water use resulted in a decrease in conductance normalized by leaf area. Water use efficiency has increased slightly since 2008 with an average of 10.0 g CO2 kg−1 H2O for the last 6 years. No prolonged dry periods were observed during the study period. The functioning of this novel ecosystem is evolving as expected on the basis of the trends observed for other natural and disturbed boreal forests.