Interactive comment on “ Impacts of droughts on carbon sequestration by China ’ s terrestrial ecosystems from 2000 to 2011 ” by Y . B

In recent years, China's terrestrial ecosystems have experienced frequent droughts. How these droughts have affected carbon sequestration by the terrestrial ecosystems is still unclear. In this study, the process-based Boreal Ecosystem Productivity Simulator (BEPS) model, driven by remotely sensed vegetation parameters, was employed to assess the effects of droughts on net ecosystem productivity (NEP) of terrestrial ecosystems in China from 2000 to 2011. Droughts of differing severity, as indicated by a standard precipitation index (SPI), hit terrestrial ecosystems in China extensively in 2001, 2006, 2009, and 2011. The national total annual NEP exhibited the slight decline of −11.3 Tg C yr −2 during the aforementioned years of extensive droughts. The NEP reduction ranged from 61.1 Tg C yr −1 to 168.8 Tg C yr −1 . National and regional total NEP anomalies were correlated with the annual mean SPI, especially in Northwest China, North China, Central China, and Southwest China. The reductions in annual NEP in 2001 and 2011 might have been caused by a larger decrease in annual gross primary productivity (GPP) than in annual ecosystem respiration (ER). The reductions experienced in 2009 might be due to a decrease in annual GPP and an increase in annual ER, while reductions in 2006 could stem from a larger increase in ER than in GPP. The effects of droughts on NEP lagged up to 3–6 months, due to different responses of GPP and ER. In eastern China, where is humid and warm, droughts have predominant and short-term lagged influences on NEP. In western regions, cold and arid, the drought effects on NEP were relatively weaker but prone to lasting longer.


Introduction
The exchange of CO 2 between land and the atmosphere is an important pathway in the global carbon cycle (Arnone et al., 2008;Schwalm et al., 2010) change.Extreme climate events, such as drought, could substantially affect ecosystem carbon fluxes and lead to significant inter-annual variability in regional and even global terrestrial carbon budget (Running, 2008;Xiao et al., 2011).In contrast to the gradual climate change, drought might intensively impact carbon sequestration by terrestrial ecosystems in a relatively short period through strong influences on both gross primary productivity (GPP) and respiration (RE), which releases carbon to the atmosphere (Baldocchi, 2005;Meir et al., 2008;Schwalm et al., 2010;van der Molen et al., 2011).More frequent and severe droughts were projected to occur under future climate change, characterized by decreasing precipitation and increasing temperature, especially in the middle and high latitudes (IPCC, 2007;Dai, 2013).Owing to strong coupling between the carbon and water cycles (Law et al., 2002), the effects of drought events on terrestrial carbon sequestration have been intensively investigated in recent years (Cox et al., 2000;Zeng et al., 2004;Ciais et al., 2005;Wu and Chen, 2013).Models and remote sensing have been acted as effective tools for assessing the response of terrestrial carbon cycle to drought at regional and global scales (Chen et al., 2012a).Several modeling studies have been recently conducted to assess the impacts of drought on the productivity and carbon budget of terrestrial ecosystems for western North America (Schwalm et al., 2012), the southern United States (Chen et al., 2012a), Europe (Ciais et al., 2005;Reichstein et al., 2007;Vetter et al., 2008), Amazonia (Phillips et al., 2009;Potter et al., 2011;Nunes et al., 2012), East Asia (Saigusa et al., 2010;Xiao et al., 2009), the Northern Hemisphere mid-latitudes (Zeng et al., 2005), and the globe (Zhao and Running, 2010;Chen et al., 2013).All studies indicated that droughts significantly influenced the terrestrial carbon cycle and might even cause regional and global terrestrial ecosystems to shift from carbon sinks to carbon sources (Chen et al., 2013;Ponce Campos et al., 2013;Piao et al., 2013).
China has the third largest land area in the world, with diverse climates and biomes (Cao et al., 2003) carbon budget in China is of great significance for studying past, current and future roles of terrestrial ecosystems in retarding the increase of atmospheric CO 2 concentration and climate change (Mu et al., 2008;Piao et al., 2009a;Tian et al., 2011b).The ability of terrestrial ecosystems in China to absorb carbon is limited by water availability, especially in vast arid and semiarid regions (Hu et al., 2010;Yi et al., 2010;Yu et al., 2013).Many severe drought episodes have occurred in China since the 1950s (Zou et al., 2005;Zhai et al., 2010), and especially in the first 10 yr of this century (Piao et al., 2009b;Qin et al., 2010;Lu et al., 2011;Wang et al., 2011;Wu et al., 2011).These droughts significantly influenced terrestrial carbon cycle at regional and national scales (Xiao et al., 2009).
Several studies have been devoted to investigate the impacts of droughts on regional or national terrestrial carbon budget in China.Working with tree-ring width chronologies and the Terrestrial Ecosystem Model (TEM), Xiao et al. (2009) discovered that severe and extended droughts during the 20th century substantially reduced ecosystem carbon sequestration, or even switched terrestrial ecosystems from carbon sinks to carbon sources.Using the Carnegie-Ames-Stanford approach (CASA) model, Pei et al. (2013) declared that some droughts in China during the period from 2001 to 2010 substantially reduced the countrywide net primary productivity (NPP), whereas others did not.Utilizing the MODIS GPP and NPP products, Zhang et al. (2012c)  2 Data and methods

Conclusions References
Tables Figures

Back Close
Full

NEP calculation
NEP is calculated as the difference between photosynthesis and respiration: where R h is the heterotrophic respiration, and maintenance respiration, R m , and growth respiration, R g , are the components of the autotrophic respiration.R g is assumed to be 25 % of GPP and R m is calculated as a function of biomass, temperature, and reference respiratory rate at 25 • C (Ju et al., 2006;Mo et al., 2008): where M i is the biomass carbon content (kg m −2 ); subscript i denotes leaf, stem, coarse root, and fine root carbon pools, respectively; r m,i is the respiration rate at a base temperature T b , set as 15 • C; T is the temperature.The BEPS model simulates the dynamics of soil carbon using the methodology borrowed from the Century model (Parton et al., 1993).R h is the total of carbon released from five litter carbon pools: surface structural, surface metabolic, soil structural, soil metabolic, and coarse woody and from four soil carbon pools: surface and soil microbes, slow, and passive (Ju et al., 2006;Mo et al., 2008): where τ j is the prescribed respiration coefficient of pool j; k j the decomposition rate of pool j which equals the prescribed maximum decomposition rate down-scaled according to soil temperature, water content, texture, and lignin content; and C j is the size of pool j, updated for each time step.Introduction

Conclusions References
Tables Figures

Back Close
Full

Regulation of stomatal conductance by soil water content
In the BEPS, the Jarvis model used to determine stomatal conductance of sunlit and shaded leaves (Jarvis, 1976) is regulated by temperature, atmospheric vapor pressure deficit, photosynthetic photon flux density, and soil water content.Assuming that vegetation is able to optimize the uptake of soil water, the soil water scalar, f w , is determined from: where f w,i is the soil water stress factor in layer i, and β i is the weight of layer i expressed as a function of soil water availability and root abundance (Ju et al., 2006): where r i is the root fraction within layer i as determined from Zhang and Wegehenkel. (2006).The term f w,i in layer i is calculated as a function of volumetric soil water content, wilting point, field capacity, and porosity of soil layer i (Chen et al., 2005;Ju et al., 2010a).

Data used
The BEPS model is driven by both spatially variant and invariant datasets.Atmospheric CO 2 concentration is assumed to be spatially homogenous.The spatially variant inputs into the BEPS include: (1) Remote sensing data: these include yearly MODIS land cover datasets (MCD12Q1 V051) from 2001 to 2010 (Friedl et al., 2010)  reflectance product (MOD09A1 V05) and an algorithm based on a 4-scale geometric optical model (Deng et al., 2006;Liu et al., 2012).Previous studies proved that LAI inverted using this algorithm has superiority over the existing MODIS product (Pisek et al., 2007;Garrigues et al., 2008).
(2) Meteorological data: daily maximum and minimum air temperatures, precipitation, relative humidity, and incoming solar radiation required to force the BEPS model were interpolated from observations at 753 basic meteorological stations across China using the inverse distance weight method.In the interpolation of temperatures, a lapse rate of 6 • C per 1000 m was assumed.Incoming solar radiation is not observed at all meteorological stations, so it was estimated according to sunshine duration measurements at stations without solar radiation observations.
(3) Soil data: the volumetric fractions of clay, sand, and silt were interpolated from the soil texture maps developed by Shangguan et al. (2012) on the basis of the 1 : 1 000 000 scale soil map of China and 8595 soil profiles recorded in the second national soil survey dataset.The fractions of clay, sand and silt were used to estimate hydrological parameters, including the wilting point (water potential at 1500 kPa), field capacity (water potential at 33 kPa), porosity, saturated hydrological conductivity, and air entry water potential.

Drought assessment
A variety of indices have been developed to assess drought severity (Mishra and Singh, 2010;Dai, 2011).In this study, SPI was used as the indictor of drought severity, owing to its simplicity, temporal flexibility and spatial consistency (Hayes et al., 1999;Chen et al., 2012a).Recently, many studies have proved SPI to be effective at monitoring drought in China (Zhai et al., 2010;He et al., 2011;Pei et al., 2013;Zhang and Jia, 2013).The 1, 3, 6, 9 and 12 months SPIs for each month were calculated using monthly precipitation data over the period from 1970 to 2011.For a given month, the L (L = 1, 3, 6, 9, 12) months SPI is calculated using monthly precipitation data in the current and Introduction

Conclusions References
Tables Figures

Back Close
Full previous (L − 1) months.Following McKee et al. (1993) and the classification standard of meteorological drought issued by the China Meteorological Administration (Zhang et al., 2011), precipitation deficit in a given year was classified into 9 categories according to the 12 month SPIs through the end of December (Table 1).

Assessment of accumulative lagged effects of drought on carbon sequestration
The accumulative lagged effects of drought on carbon sequestration was assessed on the basis of the correlations between monthly NEP SAI values and monthly SPI.For each pixel, the correlation coefficient can be calculated as follows: R(m, L) = For better analysis of spatial and temporal patterns of drought events and of the terrestrial carbon cycle responses, China was divided into nine climatic zones (Piao et al., 2009a), which are displayed in Fig. 1.

Spatial and temporal variations of drought in China during 2000-2011
Figure 2 shows the national means of annual SPI (the 12 month SPIs through the end of December) in China for the period from 1970 to 2011, indicating more droughts occurred in China from 2000 to 2011, in comparison with previous three decades.The national annual mean SPI was below zero in 2000, 2001, 2004, 2006, 2009 and 2011.It was even smaller than −0.20 in 2001, 2006, 2009 and 2011, indicating that China was widely hit by severe droughts in these four years.In 2001 and 2009, areas hit by different degrees of droughts accounted for 46.9 % and 48.8 % of the national total, respectively.Droughts in 2001 and 2011 were more serious, with above 20 % of the national total area affected by moderate, severe, and extreme droughts (Fig. 3).In 2001In , 2006In , 2009In and 2011, the differences in the severity and areas of drought were distinguishable.In 2001, drought occurred in all regions but the Tibetan Plateau, South and Southwest China (Fig. 4).Areas affected by moderate and severe droughts were mainly located in Northeast China, Inner Mongolia, North China, and Central China.The drought in 2011 was recognized as the most serious one in the recent 60 yr (Sun and Yang, 2012).Vast areas of Southwest, Central, South and Southeast China Introduction

Conclusions References
Tables Figures

Back Close
Full were hit by moderate to extreme droughts, and those in Yunnan, Guizhou and Hunan Provinces were extremely serious.Drought also occurred widely over China in 2009.However, the severity of drought was mostly mild to moderate (Gao and Yang, 2009;Lu et al., 2011;Barriopedro et al., 2012).Moderate drought mainly occurred in the south parts of both Southwest China and the Tibetan Plateau.In 2006, drought was mostly mild (Fig. 3).Moderate drought only occurred in areas around Chongqing (Wang et al., 2011;Wu et al., 2011).

Response of carbon sequestration to droughts
Figure 5 shows the annual departures of national total NEP from the multiyear mean over the period from 2000 to 2011.During the past 12 yr, the national total of NEP exhibited a slightly decreasing trend, at an average rate of −11.3 Tg C yr −2 .This decline of national NEP was mainly caused by a significant decrease of NEP in 2001NEP in , 2006NEP in , 2009NEP in and 2011.In these four typical drought years, the national totals of NEP respectively decreased by 62.9, 88.0, 168.In 2001, the decrease of simulated NEP mainly occurred in regions north of the Yangtze River (the blue and cyan colored areas in Fig. 6), including most of North China, the north parts of Southeast and Central China, and the east of Northwest China (Fig. 6).In drought affected regions, both annual GPP and RE decreased owing to low water availability.The reduction of GPP was larger than that of RE in most areas, Introduction

Conclusions References
Tables Figures

Back Close
Full resulting in decreases of NEP, which were significant in North and Northwest China.
The SAI values of total NEP in these two regions were below −1.5 (Fig. 7).
The decrease of NEP in 2006 mainly occurred in the central areas of China (the blue and cyan colored areas in Fig. 6), such as the north parts of Southwest and Central China, the north and middle parts of Southeast China, and the east of Tibetan Plateau.
The reduction of carbon sequestration in drought affected areas was mainly caused by the slight decrease of GPP and slight to severe increase of RE.The SAI values of regional total NEP in Southwest and Central China were −1.25 and −1.26, respectively (Figs. 6 and 7), indicating moderate reductions of regional total NEP here.
NEP decreased in most regions of China in 2009 with the exception of Northeast China, most of the Tibetan Plateau, and the western part of Northwest China (Fig. 6).The decrease of NEP was mostly caused by slight to moderate decreases of GPP and slight to severe increases in RE.Regional NEP totals decreased extremely in south China (NEP SAI = −2.23),severely in the Tibetan Plateau (NEP SAI = −1.50)and Southwest China (NEP SAI = −1.80),and moderately in Central (NEP SAI = −1.42)and Southeast China (NEP SAI = −1.08)(Fig. 7).
In 2011, moderate to extreme droughts hit regions south of the Yangtze River (Fig. 4), resulting in detectable decreases in GPP.Meanwhile, RE also decreased to some extent due to the reduction of heterotrophic respiration caused by low soil water availability.As a consequence, the decrease of NEP was smaller in 2011 than that in 2009.
The decrease of NEP mainly occurred in Central China, the eastern part of Southwest China, and adjacent areas of North and Southeast China (Fig. 6).
Figure 7 shows the anomalies of annual total GPP, RE and NEP, along with annual mean SPI in different regions of China.The anomalies of annual total NEP were positively correlated with annual mean SPI at the 0.01 significance level in Northwest, North, Central, and Southwest China.However, the linkages of NEP anomalies with annual mean SPI were not significant in Northeast China, Inner Mongolia, and the Tibetan Plateau, where temperature is the dominant factor controlling carbon sequestration.It should be kept in mind that most areas of Inner Mongolia are arid and semiarid, so veg-Introduction

Conclusions References
Tables Figures

Back Close
Full etation activity is mainly controlled by water availability.However, in the northeastern part of Inner Mongolia, the climate is humid and water availability does not act as the crucial limiting factor of vegetation activity.The large fraction of total carbon sequestration capacity in Inner Mongolia is contributed by terrestrial ecosystems in the northeast part.Therefore, total carbon sequestration capacity in Inner Mongolia is not sensitive to precipitation anomalies.In humid Southeast and South China, the total NEP anomaly is also not significantly correlated with annual mean SPI.In these regions, temperature and water availability are normally suitable for vegetation growth.Radiation is a key factor driving the interannual variability of productivity (Nemani et al., 2003).Precipitation increase is normally associated with the decrease of radiation, possibly causing NEP decrease to some extent.The simulations indicated that drought events in China during the past 12 yr generally caused detectable reductions of GPP and NEP in strongly drought-affected areas (Figs. 6 and 7).Similar negative effects of drought on carbon sequestration were also reported in some previous studies conducted in Europe (Ciais et al., 2005;Vetter et al., 2008) and America (Xiao et al., 2011;Chen et al., 2012a;Schwalm et al., 2012).

The accumulative lagged effect of drought on NEP
Figure 8 shows the maximum correlation coefficient between SPI and monthly NEP SAI for the period 2000-2011 and the SPI time-scales at which the maximum correlation between SPI and NEP SAI is found.NEP SAI is correlated with SPI at the 0.05 significance level over 82 % of the vegetated land areas (Fig. 8a), especially in Southwest, Southeast, and Northeast China.The SPI time-scales, at which correlation between SPI and NEP SAI maximizes, exhibit distinguishable spatial patterns (Fig. 8b).Drought had relatively shorter time-scales accumulative lagged effect on NEP in the southern regions of China (e.g., 1-3 month).In cropland regions, NEP responded to drought at middle time-scales (about 6 months).Drought might impact on NEP at longer time scales in cold grassland areas than in other regions.Introduction

Conclusions References
Tables Figures

Back Close
Full

The impact of drought on the national terrestrial carbon budget
Many studies indicated that terrestrial ecosystems had been acting as a carbon sink in recent decades, although estimates are still inconsistent (Cao et al., 2003;Houghton and Hackler, 2003;Pan et al., 2004;Fang et al., 2007;Piao et al., 2009a;Tian et al., 2011a).Piao et al. (2009a) suggested that terrestrial carbon sinks amounted to 0.19-0.26Pg C during the 1980s and 1990s, accounting for 28-37 % of concurrent CO 2 emissions in China.In recent years, anthropogenic carbon emissions increased very quickly in China (Wang et al., 2012a), implying the increasing importance of maintaining terrestrial carbon sinks.Our simulation indicates that the national total NEP of China slightly declined during the period from 2000 to 2011 (−11.3Tg C yr −2 ), mainly due to considerable reductions of NEP caused by widely distributed severe droughts in 2001, 2006, 2009 and 2011.In these four years, the decreases of national total NEP ranged from 61.1 Tg C yr −1 to 168.8 Tg C yr −1 , approximately 30 % to 94 % of the average carbon sink strength of forest during 1999-2008 estimated based on forest inventory data (Pan et al., 2011;Guo et al., 2013;Zhang et al., 2013).Therefore, drought is a dangerous threat to the terrestrial carbon sink in China.With possible increases in drought frequency and severity in future (IPCC, 2007;Dai, 2013), the role of China's terrestrial ecosystems in offsetting fossil fuel emission might be weakened.

The lagged effect of drought on carbon sequestration by terrestrial ecosystems
It was found here that drought has lagged effects on NEP.The time-scales, at which NEP most significantly respond to drought, spatially differ, rapid in eastern warm humid regions and slow in northern and western cold semiarid grassland regions.Similar findings have also been reported in other studies, which used greenness index, chronologies of ring width, and NPP as vegetation growth indictors for assessing the Introduction

Conclusions References
Tables Figures

Back Close
Full influences of drought on vegetation.The lag time of drought affecting growth indicators is related to vegetation types, and local conditions.Ji and Peters (2003) found out that the 3 months SPI was most strongly correlated with the growing-season NDVI in American Great Plains.Based on the analysis of NDVI-SPI co-variability over different temporal-scales, Lotsch et al. (2003) suggested that 5-month SPI be used to assess the response of terrestrial ecosystems to drought.Jain et al. ( 2010) found that the NDVI has the highest correlation with 1-month, 3-month, and 9-month SPI in different regions of India.In Texas of US, the growing-season vegetation condition index (NDVI-based) was found to be most strongly correlated with the 6-month and 9-month SPI (Quiring and Ganesh, 2010).The correlation between chronologies of ring width and different time scales SPI for eight tree species in north-eastern Spain indicated that the tree growth most significantly responded to 9-11 months SPI for xeric sites and to shorter than 5 months SPI for mesic sites, respectively (Pasho et al., 2011).
Recent studies on the effects of drought on NPP in China found out that correlations between drought intensity and NPP anomalies were the strongest during or after the drought intensity peaked (Chen et al., 2013;Pei et al., 2013).With vegetation indices from satellite imagery, tree-ring growth series, and aboveground net primary production records, Vicente-Serrano et al. (2013) recently reported the responses of global biomes to drought, both arid and humid biomes respond to drought at short time-scales while semiarid and subhumid biomes respond to drought at long time-scales.This is similar not decrease or would even increase.With the continuation of drought, water in deep soil layers will diminish and large decreases of GPP will occur, resulting in a reduction of NEP (Ju et al., 2006;Barr et al., 2007).These phenomena have been confirmed by eddy covariance measurements (Granier et al., 2007;Kljun et al., 2007).Krishnan et al. (2006) reported that after experiencing a 3 yr long drought, a boreal aspen stand in central Saskatchewan, Canada, declined in both growth and leaf area index even 2 yr after the drought.Pereira et al. (2007) demonstrated that the GPP, RE, and NEP of a Eucalyptus plantation located in southern Portugal, decreased during the drought years, the largest absolute decrease occurring in the second drought year.

Uncertainties and remaining issues
In this study, the response of terrestrial carbon fluxes in China to droughts was explored with the BEPS model, driven by remote sensing data.It should be kept in mind that there might be some uncertainties.Explicit soil water content dynamics is pivotal to understanding the impacts of drought on carbon sequestration by terrestrial ecosystems (Reichstein et al., 2002;Samanta et al., 2011).Similarly to most of other ecological models, the BEPS model used here does not account for the horizontal movement of soil water.This simplification would definitely over or underestimate the impact of drought on carbon sequestration in low or upper slopes areas.Irrigation is popularly practiced for most croplands in China, and was not included in the BEPS model.Consequently, the negative effects of drought on carbon sequestration by croplands might be overestimated.
The LAI is a key input to the BEPS model.In this study, the LAI was inverted using the MODIS reflectance data.Although the advantages of this LAI inversion algorithm have been proved in a series of previous studies, the inverted LAI still contain some uncertainties.The quality of the LAI depends largely on the quality of MODIS reflectance data, which is not guaranteed in many areas of China, especially in summer.A locally adjusted cubic-spline capping (LACC) method (Chen et al., 2006)  atmospheric noise.However, it is impossible to remove noise using the LACC method if good quality MODIS data are not available for a length period.The uncertainty in LAI would be propagated into estimated NEP. NEP is affected by a number of factors, such as climate, disturbances and land cover change.Here the focused is only on the influence of extensively severe droughts on the terrestrial NEP cycle in China.Droughts might invoke other disturbances, such as fire, disease, insects, mortality or re-growth (Allen et al., 2010;Peng et al., 2011;van der Molen et al., 2011;Ma et al., 2012).These indirect effects of droughts on terrestrial carbon sequestration were not explored here.Land cover and land use change could also significantly modify the terrestrial carbon cycle (Piao et al., 2007;Xiao et al., 2009).During the study period, both the dramatic urbanization (Wang et al., 2012b) and large-scale forest plantation programs have concurrently occurred in China (Piao et al., 2012), thus impacting terrestrial carbon cycles at regional and national scales.These effects were not explicitly quantified here and require future research.

Conclusions
In this study, the influences of drought on carbon sequestration by terrestrial ecosystems during the period from 2000 to 2011 were assessed by model simulation and remote sensing data.The main conclusions are: anomalies in Northwest, North, Central, and Southwest China, and not sensitive to precipitation anomalies in other regions.
(3) The drivers of national total NEP reductions differed among years.In 2001 and 2011, droughts caused both national totals of GPP and RE to decrease.The former decreased to a large degree, resulting in decreases of national total NEP.
In 2006, droughts were mostly mild or moderate, and the decrease of NEP was caused by a smaller increase of GPP compared with RE.In 2009, the largest decreases of national total NEP during the study period was caused by the decrease of GPP and increase of RE.
(4) The drought has accumulative lagged effects on carbon sequestration.The speed and degree, at which NEP responds to drought, varies spatially, rapidly and significantly in east humid and warm regions of China.In contrast, NEP responds to drought relatively slowly and less significantly in northern and western cold arid and semiarid regions.Introduction

Conclusions References
Tables Figures

BGD Introduction Conclusions References
Tables Figures

BGD BGD Introduction Conclusions References
Tables Figures

Back Close
Full Carbon sequestration by terrestrial ecosystems is affected by a number of factors, including climate, atmospheric CO 2 concentration, nitrogen deposition, land cover types, and land use Discussion Paper | Discussion Paper | Discussion Paper | . Terrestrial ecosystems in China play an important role in the global terrestrial carbon sink.Quantifying the spatial and temporal variations of the terrestrial Discussion Paper | Discussion Paper | Discussion Paper | reported that the 2010 spring drought caused detectable reductions of GPP and NPP in southwestern China.Using three light use efficiency models (CASA, MODIS-GPP and EC-LUE (Eddy Covariance Light Use Efficiency) and one dynamic vegetation model (Integrated Biosphere Simulator, IBIS), Yuan et al. (2013) found that the carbon sink over North China declined 0.011 Pg C yr −1 during the 1999-2011 drought period.However, these studies mainly focused on the influences of droughts on GPP or NPP.How the national total net exchange of carbon between the atmosphere and terrestrial ecosystems is affected by droughts has not been thoroughly explored.This study is devoted to assessing the impact of droughts on net ecosystem productivity (NEP) of terrestrial ecosystems in China over the period from 2000 to 2011 Discussion Paper | Discussion Paper | Discussion Paper | with a process-based Boreal Ecosystem Productivity Simulator (BEPS) model, which is driven by remotely sensed vegetation parameters.The objectives of this paper are to: (1) characterize the spatial and temporal variations of droughts during the period from 2000 to 2011 based on standard precipitation index (SPI); (2) evaluate the response of NEP to droughts for whole country and different regions of China; (3) assess the accumulative and lag effects of drought on carbon sequestration.
Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | and 8 day leaf area index (LAI) from 2000 to 2011.The LAI dataset were inverted using the MODIS Discussion Paper | Discussion Paper | Discussion Paper | Screen / Esc Printer-friendly Version Interactive Discussion Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | 12 n=1 NEP SAI (m, n) − NEP SAI (m) SPI(m, L, n) − SPI(m, L) 12 n=1 NEP SAI (m, n) − NEP SAI (m) 2 12 n=1 SPI(m, L, n) − SPI(m, L) Discussion Paper | Discussion Paper | Discussion Paper | where R(m, L) is the correlation coefficient between NEP SAI and SPI in month m at a given L value (L = 1, 3, 6, 9, 12) for a specific pixel, NEP SAI (m, n) is the NEP SAI value in month m of year n, NEP SAI (m) is the average of NEP SAI (m, n) over 12 yr, SPI(m, L, n) is the SPI value for a given L in month m of year n, SPI(m, L) is the average of SPI(m, L) values over 12 yr.A value of L producing the largest R(m, L) indicates that drought has accumulative lagged effects up to L months on NEP.
Discussion Paper | Discussion Paper | Discussion Paper | 8 and 61.1 Tg C yr −1 .The annual anomalies of national total NEP were correlated with national means of annual SPI (r = 0.54, p = 0.06).The main drivers for NEP reductions differed in different years.Annual total GPP and RE both decreased in 2001 and 2011 in comparison to 12 yr means.The larger decrease of GPP than RE resulted in sizeable reductions of annual total NEP in these two years.In 2006, national total GPP increased slightly while national total RE increased at a larger magnitude, causing notable decrease in the national NEP.The drought in 2009 caused a decrease in national total GPP and an increase in national total RE, resulting in the largest decrease of national total NEP in the study period.
Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | to our findings in this study.Based on in situ eddy covariance flux data and meteorological observation data at Qianyanzhou station, southern China, Huang et al. (2013) found drought exerted an approximate one month negatively lagged effect on potential productivity.NEP is the balance between GPP and RE.The response of NEP to drought depends on how GPP and RE react to drought.At the initiation stage of drought, surface soil becomes dry but deep soil is still wet, so plants are able to take in water from deep soils to maintain normal stomata functional and carbon assimilation.Meanwhile, heterotrophic respiration is limited by low water availability in surface soil.Consequently, NEP would Discussion Paper | Discussion Paper | Discussion Paper | was employed here to remove unrealistic fluctuations of LAI caused by residual cloud contamination and Discussion Paper | Discussion Paper | Discussion Paper |

( 1 )
Drought, as indicated by the SPI, occurred extensively and frequently in China during the period from 2000 to 2011, especially in 2001, 2006, 2009 and 2011.The spatial patterns of the droughts differed in these four years.(2) The national total NEP slightly declined by −11.3 Tg C yr −2 during the period from 2000 to 2011, mainly due to the large reductions of NEP in 2001, 2006, 2009 and 2011.Anomalies of national total NEP were correlated with annual SPI (r = 0.54, p = 0.06).The regional totals of annual NEP were sensitive to precipitation Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Ciais, P., Discussion Paper | Discussion Paper | Discussion Paper | intercomparison of global Leaf Area Index products derived from remote sensing data, J.