Changes and determining factors of crop evapotranspiration derived from satellite-based dual crop coefficients in North China Plain

Qinghua Tan ,Yujie Liu, *,Tao Pan,* ,Xianfang Song ,Xiaoyan Li

a Key Laboratory of Land Surface Pattern and Simulation,Institute of Geographic Sciences and Natural Resources Research,Chinese Academy of Sciences,Beijing 100101,China

b University of Chinese Academy of Sciences,Beijing 100049,China

c Key Laboratory of Water Cycle and Related Land Surface Processes,Institute of Geographic Sciences and Natural Resources Research,Chinese Academy of Sciences,Beijing 100101,China

d State Key Laboratory of Earth Surface Processes and Resource Ecology,Faculty of Geographical Science,Beijing Normal University,Beijing 100875,China

Keywords:Actual crop evapotranspiration Determining factor North China Plain Rotation system Spatiotemporal variation

ABSTRACT Evaluating actual crop evapotranspiration(ETc) variations and their determining factors under changing climates is crucial for agricultural irrigation management and crop productivity improvement in nonhumid regions.This study analyzed the spatiotemporal characteristics and detected the determining factors of ETc for winter wheat and summer maize rotation system from 2000 to 2017 in the North China Plain (NCP),by combining the FAO-56 dual crop coefficient approach with remotely sensed vegetation indices (VIs).The results indicated that daily air temperature increased in varying degrees while wind speed and sunshine hours decreased slightly during the growing season of winter wheat and summer maize over the study period.The trends of relative humidity and effective precipitation varied in crop growing seasons.Based on the validated relationship of dual crop coefficients and VIs,the estimated multi-year average ETc of winter wheat (370.29 ± 31.28 mm) was much higher than summer maize(281.85 ± 20.14 mm),and the rotation cycle was 652.43 ± 27.67 mm.Annual ETc of winter wheat and the rotation cycle increased by 2.96 mm a-1 and 1.77 mm a-1,respectively.However,the ETc of summer maize decreased with distinct spatial variation.Spatially,winter wheat ETc increased significantly in the northeast NCP,covering the Beijing-Tianjin-Hebei areas.Meanwhile,significant increases in summer maize ETc were detected in the southwest NCP.The sensitivity and contribution analysis showed that ETc of winter wheat and summer maize was positively sensitive to temperature,wind speed,and sunshine hours while negatively to relative humidity.Moreover,wind speed and sunshine hours contributed most to changes in ETc (around 20%-40%).

1.Introduction

Crop evapotranspiration(ETc),reflecting the crop’s water needs imitated by transpiration and evaporation,is a vital component of describing the hydrological cycle in ecological systems,estimating water balance and scheduling irrigation [1,2].Climate change,characterized by rising air temperature and uneven precipitation,intensifies the changes in crop ETcand enhances stress on the agricultural system[3,4].Investigation of the variation of ETcunder climate change is critical for optimizing irrigation management and improving crop yield [5,6].

The ETccan be measured using a lysimeter,eddy covariance,Bowen ratio,and soil water balance[7,8].However,these methods cannot determine the distributed or regional scale evapotranspiration[9-11].Empirical equations and models were developed based on the agronomic,biophysical,and meteorological elements as input variables.Due to its simplicity and effectiveness,FAO-56 single and dual crop coefficient approaches were widely used to estimate ETcby multiplying the crop coefficient(Kc)with the reference evapotranspiration(ET0)[12,13].The Kcconsists of basal crop coefficient (Kcb) and soil evaporation coefficient (Ke).Determined by crop type,climate,tillage practices,soil water management,fertilization,and plant growth,Kcis dynamic during the crop growing season.Local adjustments for stage duration and basal crop coefficients are expected to be more suitable and accurate for estimating ETcand crop water demand than the use of tabulated values[13,14].Remotely sensed vegetation index (VIs) data can reflect the crops’actual growth stage and status due to its improved temporal coverage and spatial resolution,and has already been used to predict actual ETcover large agricultural fields [13-15].The wellestablished relationships of Kcb/Kcand remotely sensed VIs such as Normalized Difference Vegetation Index(NDVI)had assessed ETcin a simple,robust,and operative manner [16,17].Thus,the remote sensing approach enables timely estimation of crop water use for resource monitoring and irrigation scheduling [18,19].Moreover,variations in weather conditions affect crop water-use patterns,resulting in an inaccurate estimation of ETc[19,20].A warmer climate may increase the potential evapotranspiration of crops,leading to greater demand for irrigation water [21].In addition,other factors such as sunshine hours,wind speed,and relative humidity also play important roles in the variation of ETc.The impact of climate change on ETcwas widely evaluated,but various or even contrary results have been concluded in different regions [10,22-24].For example,Jia et al.[22]indicated that decreased solar radiation,wind speed,and sunshine duration resulted in a decline in ETcover the past five decades in Northeast China.By contrast,changes in sunshine duration,air temperature,and wind speed contributed negatively to the increase of total ETcin the loess plateau of China[25].Analyzing the sensitivity of ETcto climate change and identifying the key climate factors can help understand the characteristic of ETcvariation and formulate reasonable irrigation schemes.

The North China Plain (NCP) is one of the granaries in China,where winter wheat and summer maize are predominant crops.Irrigation is widely used to improve crop production and resource use efficiency due to a mismatch between crop water requirements and precipitation during the winter wheat-summer maize growing season[6].Approximately 75% of the agricultural land is irrigated,consuming 70%-80% of the total water resource allocation in the NCP [26,27].Overall,agricultural water consumption mainly consisted of groundwater irrigation,one of the biggest challenges to agriculture sustainability [28,29].Studies on crop ETcmeasurement have been conducted in this region [30,31],in which the FAO-56 crop coefficient approach was mainly applied [32,33].However,the crop coefficient at different growth stages was kept constant or adjusted using tabulated methods recommended by FAO in most studies [34].Although a few studies evaluated the Kc-VIs relationship at the station scale,the dynamic characteristics and regional variations of the crop coefficients were not fully captured in the NCP [8,35].Moreover,the soil evaporation coefficient was overlooked in some studies,leading to uncertainties in ETcestimation.A comprehensive assessment of crop evapotranspiration based on dynamic dual crop coefficients has not yet been provided regionally.Further,significant climate change such as climate warming and decreasing solar radiation has been observed in the NCP in recent decades,affecting agricultural water use and crop yield [36].Research on the impact of climate change on ETcin the NCP is still insufficient.It is necessary to analyze the spatiotemporal variation of ETcand detect its determining factors using dynamic crop coefficients in the NCP.

Therefore,by combining remotely-sensed VIs data and the FAO-56 dual crop coefficient approach,the objectives of this study are to(1)clarify the variations in climatic factors during crop growing seasons from 2000 to 2017,and (2) evaluate the spatiotemporal changes of ETcfor the winter wheat-summer maize rotation system,and (3) investigate the responses of ETcto climate change and detect the determining factors in the NCP.

2.Materials and methods

2.1.Study area and experimental stations

The NCP is located in eastern China,covering the entire Beijing and Tianjin municipalities,the majority of Hebei,Shandong,and Henan provinces,and the northern parts of Jiangsu and Anhui provinces.The NCP is an alluvial plain developed by the intermittent flooding of the Huang (the Yellow River),Huai,and Hai rivers,and 85% of the land is cultivated as farmland.Additionally,the NCP has an average elevation of 73.4 m.The region is dominated by a temperate monsoon climate,with a mean annual temperature of 8-15°C.The average annual precipitation is 500-1000 mm,and approximately 70%occurs from June to September[37].The double cropping system of winter wheat-summer maize is the most common cultivation pattern in NCP.The harvested areas of winter wheat and summer maize in 2015 (Fig.1) were extracted from Grogan et al.[38].Generally,winter wheat is sown in mid-October and harvested in early June of the following year.Summer maize is sown in mid-June and harvested in early October.Due to insufficient precipitation in the growing season and irregular distribution among seasons,irrigation is essential to maintain crop growth and enhance yield.

Three comprehensive experimental stations,including Yucheng(116°38′E,36°57′N),Luancheng (114°41′E,37°53′N),and Fengqiu(114°24′E,33°01′N),are distributed in the central,southern and northern parts of the NCP,respectively (Fig.1).The soil texture of loam and sandy loam in the three stations is similar and representative of the NCP.Field experiments were performed,and winter wheat and summer maize were planted continuously in the lysimeters.The structure,functions,and principles of the lysimeter were described by Yang et al.[39].Soil water moisture was measured once every-five days,and additional measurements were made before and after irrigation and after every rainfall.More details about the experimental sites can be found in previous research [32,34,35].In addition,flood irrigation was implemented during the crop growing season.The soil moisture status of the two crops in the Yucheng station was analyzed in early works[32].This study showed that winter wheat experienced little water stress,while summer maize is subject to water stress in years with less precipitation.

2.2.Data resources and preparation

The irrigation and weighing lysimeter data for winter wheat and summer maize rotation system observed at three comprehensive experimental stations were obtained.The observation years of Luancheng (2002-2006),Yucheng (2000-2005),and Fengqiu(2004-2007) were inconsistent.The phenology of winter wheat and summer maize,including sowing,maturity,and harvest dates from 2000 to 2017,was observed at the Yucheng station and applied to characterize crop growth in the NCP.Additionally,the meteorological data,including minimum and maximum temperatures,relative humidity,rainfall,wind speed at a level of 10 m h-1,and sunshine hours on a daily timescale,were collected from 2000 to 2017 at the 363 meteorological stations in the NCP (Fig.1).The daily meteorological data were calculated as cumulation or average values for the crop growing season.Moreover,the MODIS/Terra products were chosen because they have high resolution and can derive several commonly used VIs.The 16-day MODIS NDVI products for the Terra (MOD13Q1) platform with a spatial resolution of 250 m were obtained from the National Aeronautics and Space Administration’s(NASA’s) Earth Observing System Data and Information System (https://www.earthdata.nasa.gov/).

2.3.Estimation of crop evapotranspiration and adjustment for crop coefficients

FAO-56 introduced the dual crop coefficient approach,which improved the accuracy of the ETcestimate by improving the accuracy of the evaporation estimate [14].This study calculated crop coefficients Kcband Keby the FAO-56 dual crop coefficient approach at three comprehensive experimental stations.Then the relationship between VIs and crop coefficients was developed to estimate ETc.The ETcof winter wheat and summer maize were calculated as follows:

Fig.1.Study area and distribution of winter wheat (A) and summer maize (B).

where ET0is the crop reference evapotranspiration (mm d-1).FAO-56 Penman-Monteith equation is the most widely used algorithm to quantify ET0,and the ET0is computed as:

where Δ is the slope vapor pressure curve (kPa °C-1);Rnis the net radiation at the crop surface (MJ m-2d-1);G is the soil heat flux density (MJ m-2d-1);γ is a psychrometric constant (kPa°C-1);T is the mean daily air temperature at the height of 2 m(°C);u2is the wind speed at the height of 2 m(m s-1);esis the saturation vapor pressure,and eais the actual vapor pressure.

According to the FAO-56[12],the growing season of the crop is divided into four distinct growth stages,the initial,developmental,mid-growth,and end-growth stages.Following Liu and Luo [32],the length of the first three stages for winter wheat was assumed to be 130,50,and 30 days of winter wheat growth period and 20,40,and 30 days of summer maize,respectively.Kcbvalues for the initial,mid-and end-phase of winter wheat were adjusted by the meteorological data [12].

2.4.Establishment and validation of the relationship between crop coefficients and VIs

The crop coefficients Kcbat three experimental stations were obtained by following the FAO-56 dual crop coefficients approach.On this basis,the relationship between crop coefficient and VIs was established.The Kcband Kewere calculated as follows:

where Esrepresents observed soil evaporation;fcis the fraction of vegetation cover;F1and F2indicate the functions of fc.The fcis strongly related to NDVI,and the relationship between fcand NDVI is as follows:

where NDVIminof each pixel is the minimum NDVI value during the growing period.In this study,experimental data from three sites in different years were used to obtain crop coefficients using the FAO-56 dual crop coefficient method.At Yucheng station,observed data from 2000 to 2002 for winter wheat and data in 2000 and 2003 for summer maize were used.In addition,observation data of winter wheat and summer maize rotation cycle from 2002 to 2004 at Luancheng and 2004 to 2006 at Fengqiu stations were used for calculation.To be compatible with the temporal resolution of the NDVI,the 16-day average values of calculated crop coefficients at the three stations were used to establish the relationships with NDVI data at the same periods.And then,the relationship was validated by observing ETcand calculating ETcfor winter wheat (from 2002 to 2005 at Yucheng station,from 2005 to 2007 at Luancheng station,and from 2006 to 2007 at Fengqiu station) and summer maize (2005 at Yucheng station,2004 at Luancheng station and 2006 at Fengqiu station).The determination coefficient R2was used to test the performance.

2.5.Estimation of effective precipitation

Effective precipitation (Pe) is the fraction of rainfall excluding surface runoff,deep percolation,or evaporation[40].In this study,Pe during the crop growth period can be simply approximated using Eq.(6) by following the U.S.Department of Agriculture Soil Conservation Method [9,41,42]:

where P is the daily rainfall (mm d-1).

2.6.Trend analysis and significance test

A simple nonparametric procedure was applied to calculate the true magnitude of the linear trend [43].The estimation given by the Theil-Sen estimator was:

where Slope is the monotonic increase or decrease rate,or the linear slope,of the entire data series Xior any segmentation Xj.When Slope is greater than 0,it indicates that the time series present an upward trend,and when Slope is less than 0,it indicates that the variable decreases.

The Mann-Kendall (MK) test,one of the most widely accepted nonparametric tests,was applied to detect significant trends in a time series [44,45].The standardized test statistic (Z) was calculated,and it indicates a significant upward/downward trend while|Z| greater than 1.96 (at a significance level of 0.05).In addition,ttest was used to test the significance of linear regression coefficients.

The relative change (RC) of each factor was expressed as the percentage of change during the study period relative to its absolute mean value:

2.7.Sensitivity and contribution analysis

The sensitivity of actual ETcto climate change was evaluated.The relative sensitivity coefficient (S) was calculated as follows[46]:

where ΔX represents the relative change in the factor,and ΔETcrepresents the relative change in ETcinduced by ΔX.Positive(negative) S values indicate that the ETcvariation is consistent with(contrary to) the factor variation.The higher the S value,the greater the factor impact.

The changes in the ETcinduced by one factor were determined by multiplying the relative change by the corresponding S value.The ETcwas recalculated by making a ± 10% change in each factor to calculate the S value by assuming that all other factors remained unchanged.

where C represents the contribution of one factor to the ETcvariations.The contribution rate of a single factor can be represented by the percentage of the contribution induced by one factor relative to the total contribution.

3.Results

3.1.Variation in climatic factors

The temporal trends of regional average climate factors during the winter and summer maize growing season in NCP are shown in Fig.2.Across the NCP,the daily temperatures increased during both winter wheat and summer maize growing seasons.The increase in the minimum temperature during the summer maize growing season was 0.02°C per year,greater than that of the daily average and maximum temperature.However,the daily maximum temperature change rates were the highest for winter wheat.Contrary to air temperature,the sunshine hours and wind speed exhibited slight decreasing trends during the crop growing season.Moreover,the relative humidity decreased by 0.24% per year during the winter wheat growing season,but increased by 0.04% per year for summer maize.Meanwhile,the variation in sunshine hours and average temperature trends of winter wheat was greater than those for summer maize.In the linear fitting of regional average climatic factors,only wind speed during winter wheat growing seasons passed the significance test of 0.05.

Fig.2.Trends of annual meteorological factors during winter wheat and summer maize growing seasons in North China Plain.

The statistical results of climate factors change trends at all stations indicated that the spatial variation of climate during the winter wheat growing season was more distinct than that of summer maize(Table 1).From 2000 to 2017,the change rates of daily minimum temperature ranged from-0.19 to 0.20°C a-1,and the standard variation was 0.06°C a-1among all stations during the winter wheat growing season.In addition,the changes in relative humidity also displayed obvious spatial differences.

Table 1 Trends of climatic factors during winter wheat and summer maize growing season.

The effective precipitation during the growing season for winter wheat and summer maize was estimated,and its temporal changes and spatial distribution showed distinct variations(Fig.3).Overall,the annual average Pe was 105±19.85 mm,121.7±21.19 mm,and 225.18±20.84 mm for winter wheat,summer maize,and rotation cycle,respectively.As shown in Fig.3,the annual Pe increased with a trend value of 1.32 mm a-1during winter wheat growing periods while it decreased by -0.32 mm a-1in summer maize growing periods.In general,the annual Pe during the rotation cycle periods slightly increased by a change rate of 0.37 mm a-1,but the change was nonsignificant.Spatially,the multi-year average Pe graduallydecreased from the southern area to the northern part of the NCP for the winter wheat and rotation cycle (Fig.3B,D).Conversely,Pe for summer maize was high in the southern NCP while low in the middle NCP.

Fig.3.Spatial-temporal variation of effective precipitation(Pe)for winter wheat-summer maize rotation system.(A)Interannual change of Pe.(B-D)Multi-year average Pe.

3.2.Spatiotemporal characteristics of crop evapotranspiration

The relationship between crop coefficients and VIs was established and validated at three comprehensive experimental stations.The results showed that ETcvalues estimated by the VI-derivation method agree well with the observations (Fig.S1;Table S1).The dual crop coefficients were then calculated,and crop evapotranspiration was quantified from 2000 to 2017.In general,the multi-year average ETcwas 370.29±31.28 mm,281.85±20.14 mm,and 652.43 ± 27.67 mm for winter wheat,summer maize,and rotation cycle,respectively.The annual ETcfor winter wheat ranged from 248.89 mm to 326.97 mm,decreasing gradually from southern areas (Anhui and Jiangsu) to northern regions (Tianjin and Hebei)in NCP(Fig.4A,D).The ETcof summer maize and the rotation cycle displayed a similar spatial distribution (Fig.4).Moreover,the annual ETcof summer maize was higher in the east and lower in the west,with the smallest values distributed in Beijing.

During the study period,the annual ETcof winter wheat and rotation cycle increased by 2.96 mm a-1and 1.77 mm a-1,respectively(Fig.4A,B).However,it decreased with a trend of-0.46 mm a-1for summer maize (Fig.4C).The MK and Sen test results showed that the trends of annual ETcduring winter wheat and summer maize growing seasons varied greatly in the region(Fig.5).Annual ETcduring the winter wheat growing season increased in the northern and central NCP region,while it decreased in southern areas.Significant increases in winter wheat ETcwere observed in the northeast of NCP,mainly in the Beijing-Tianjin-Hebei areas,where the change rates were greater than 5 mm a-1(Fig.5A,B).By contrast,the annual ETcduring the summer maize growing season decreased in most areas of northern NCP and significantly decreased at the edge of northern NCP (at significance level of 0.05).Further,the annual summer maize ETcincreased in the middle and southwest NCP,and the areas with significant increase were sparsely distributed in the southwest regions (Fig.5C,D).

3.3.Contribution of climatic factors to ETc variation

For both winter wheat and summer maize,the ETcwas positively sensitive to the daily maximum temperature,minimum temperature,wind speed,and sunshine hours,while negatively sensitive to relative humidity (Fig.6).Regarding winter wheat,the sensitivity of ETcto sunshine hours was highest,with a mean S value of 0.33,indicating that 10% changes in sunshine hours could lead to 3.3% changes in ETc.In addition,the ETcsensitivity to daily maximum temperature (0.20) and relative humidity(-0.16) were also high.However,the lowest ETcsensitivity was toward daily minimum temperature,with S values lower than 0.05.Similarly,ETcwas also sensitive to sunshine hours and daily maximum temperature during the summer maize growing season,with high S values ranging from 0.36 to 0.54 and from 0.12 to 0.37,respectively.In particular,the variation of ETcfor summer maize was the least sensitive to the wind speed during the study period.

Fig 6.Sensitivity of the winter wheat (left) and summer maize (right) evapotranspiration to climatic factors.

The contributions of different climatic factors to variations in the ETcwere presented through the sensitivity analyses and variation trends.As shown in Table 2,the average contribution rate of the wind speed to winter wheat ETcwas 19.73%,higher than other climatic factors.In addition,the decrease in sunshine hours also contributed most to the variation of winter wheat ETc,with an average contribution rate of 11.85%.On the contrary,increased daily maximum and minimum temperatures caused an increased variation of 8.17% and 3.99% in the winter wheat ETc.However,the contribution of climatic factors to changes in summer maize ETcshowed obvious variation.The increasing sunshine hours resulted in an average increase of 12.16% in the summer maize ETc,contributing the most to change ETc.Meanwhile,the increasing daily maximum temperature contributed approximately 9.82%to the summer maize ETc.However,a decrease in wind speed was the most important factor for the significant decrease in summer maize ETc,with an average contribution rate of -7.33%.

Table 2 Average contribution rates (%) of climatic factors to variations in crop evapotranspiration (ETc) under different scenarios.

Fig.4.Temporal and spatial distribution of annual crop evapotranspiration of winter wheat(left panel),summer maize(middle panel),and rotation cycle(right panel)in the North China Plain.

4.Discussion

4.1.Variations in actual ETc and its determining factors

Understanding the spatiotemporal variation of ETcin the winter wheat-summer maize rotation system is highly conducive to better NCP irrigation scheduling and agricultural water resources management.Based on dynamic crop coefficients derived from VIs,this study showed that the multi-year average ETcof winter wheat (370.29 ± 31.28 mm) and summer maize (281.85 ± 20.14 mm) were less than the crop water requirement (412 mm of winter wheat)reported by a previous study[47].It could be attributed to the assumption of the study since the current study calculated the crop evaporation under non-standard conditions,and the crops suffered water stress to a certain degree [13].In addition,differences in the research period,data type,and study area boundary also bring discrepancies to the results [48].Moreover,the overall increases in annual ETcfor winter wheat and decreased ETcduring the summer maize growing season were consistent with previous studies[37,49].The trends of ETcduring winter wheat and summer maize seasons showed distinct spatial variation in the NCP.Notably,significant increases in ETcof winter wheat were observed in the Beijing-Tianjin-Hebei region,northeast NCP,increasing the pressure on water resources utilization.

Climate change has affected the hydrological cycle and increased stress on agricultural water resources [50].The increase in temperature during the crop growing season of both winter wheat and summer maize concluded in this study was in line with previous studies,indicating that the crops were suffering from warming in NCP [51].Warming may affect irrigation water demands in two ways: first,evapotranspiration increases due to increased radiation,rising temperature,and uneven precipitation;second,the warming climate can increase the drought risk at the key growth stages[52].It primarily occurs through increasing crop water requirements and reducing the available crop irrigation water during the growing period with earlier planting and harvest dates.Our results indicated that the sensitivity of evapotranspiration to maximum temperature was high,second to sunshine hours.However,the contribution rate of temperature was limited due to its relatively small change.In addition,contrary changes in relative humidity and reductions in wind speed and sunshine hours were observed during the winter wheat and summer maize,which resulted in the complexity of ETcchanges.Overall,the changes in wind speed contributed most to the variation in winter wheat ETc,while sunshine hours were the determining factor of summer maize ETc.Previous studies have also found that substantial changes in ET0and ETcare attributable to wind speed and solar duration(or sunshine hours)[50,53].These two factors contribute to the dynamic and thermal process of evapotranspiration.Sunshine hours supply radiation and light energy,converted to heat energy and the energy source of plant evapotranspiration.On the other hand,wind speed reduction weakens the surfaceatmosphere energy exchange [50].It is believed that a significant decrease in wind speed is the primary reason for the decline in evapotranspiration,particularly in water-deficient areas [54].However,contrasting trends of ETcwere detected for winter wheat and summer maize with decreased wind speed.It might be explained that ETcwas affected by multiple factors like climate features,plant crop characteristics,and human activities [24].

4.2.Management suggestions and uncertainty

Actual ETcis required for the assessment of regional water budget and irrigation water management decisions [55].The NCP is a global hot spot of prolonged groundwater depletion induced by irrigation agriculture [28,55].Water management and adaptation strategies are necessary to avoid water shortage and environmental deterioration caused by irrigation [27].According to irrigation observation data in three comprehensive experimental stations,three to four times irrigation of conventional flood irrigation practices were implemented during the winter wheat growing season.The ETcand effective precipitation difference was calculated to represent crop water requirement in this study.As shown in Fig.S2,additional irrigation can meet the water requirement of winter wheat and reduce water stress combining with effective precipitation.However,the results showed that the ETcof winter wheat significantly increased in the Beijing-Tianjin-Hebei regions,which increased the stress on water resource utilization.Improving irrigation efficiency and breeding new varieties with low water consumption will be effective ways to cope with water resource shortages in the area [56].Regarding rain-fed summer maize,the considerable difference between crop evapotranspiration and effective precipitation during the growing season indicated the need for irrigation schemes,especially in dry years (Fig.S2A).A recent study indicated that optimizing irrigation strategies according to precipitation categories can synchronously improve winter wheat yield and water productivity [26].It was also reported that reducing irrigation increased the water use efficiency and productivity of winter wheat-summer maize rotation [57],which could efficiently cope with water resource shortages in NCP.However,the irrigation practice experiments were mostly conducted in situ(under field conditions) or by model simulations.Real-time monitoring of water status at a regional scale should be further developed to optimize irrigation strategies.

The dynamic crop coefficients derived from VIs were considered in the current study,which provided insights for studying the impact of climate change on actual crop ETc.VI-based crop coefficients can explain variations in plant growth according to local weather conditions,site-specific differences in sowing dates and seed densities,cultivars,pests,and nutrient supply [13,18].NDVI is one of the most widely adopted indices,but its sensitivity is compromised under abnormal plant water conditions [58].Therefore,a combination of additional VIs,like the Soil Adjusted Vegetation Index,should be further used[59].Beyond the climate factors,other factors such as soil types,soil tillage,crop variety,and field management greatly affect ETc[34,60].Future studies should analyze the combined influence of environmental changes and human activities on ETc.In addition,the regional phenological characteristics of winter wheat and summer maize were assumed to be consistent with those observed at the Yucheng station.However,it varied as the planting date depended mainly on local farmers[56,61].Regional heterogeneity of crop phenology should be considered in future studies to reduce uncertainty in ETcquantification.

5.Conclusions

In this study,the spatiotemporal variation of crop ETcand its climatic determining factors were investigated through the FAO-56 dual crop coefficient approach combined with remotely sensed vegetation indices.During the growing season of winter wheat and summer maize,temperature generally increased while the wind speed and sunshine hours showed an inapparent decrease.Over the study period,ETcof winter wheat and summer maize showed distinct spatiotemporal variations in NCP.The annual ETcincreased by 2.96 mm a-1for winter wheat but decreased during the summer maize growing season with a change rate of -0.45 mm a-1.Spatially,significant elevations in the ETcof winter wheat were observed in the Beijing-Tianjin-Hebei area,and summer maize ETcwas increased in the southwest NCP.The sensitivity analysis indicated that ETcwas positively sensitive to temperature,wind speed,and sunshine hours but negatively to relative humidity.Moreover,the wind speed and sunshine hours contributed most to changes in crop ETc.Optimizing irrigation strategies and adapting more advanced irrigation and cultivation technology are required to cope with the variations in ETcunder changing climates,and to meet the urgent need for the efficient utilization and protection of water resources.The results of the study could provide support for regional water resource management and crop irrigation optimization.

CRediT authorship contribution statement

Qinghua Tan:Visualization,Writing -original draft,Writing -review &editing,Formal analysis.Yujie Liu:Conceptualization,Resources,Funding acquisition,Project administration,Writing -review &editing.Tao Pan:Conceptualization,Formal analysis,Visualization,Writing-review &editing.Xianfang Song:Writing-original draft.Xiaoyan Li:Writing -original draft.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

The research was supported by the Strategic Priority Research Program of the Chinese Academy of Sciences (XDA28060200),the National Science Fund for Excellent Young Scholars (42122003),the Strategic Priority Research Program of the Chinese Academy of Sciences (XDA20040301),and the Youth Innovation Promotion Association,CAS (Y202016).

Appendix A.Supplementary data

Supplementary data for this article can be found online at https://doi.org/10.1016/j.cj.2022.07.013.