• Allen, R. G., L. S. Pereira, D. Raes, and M. Smith, 1998: Crop evapotranspiration—Guidelines for computing crop water requirement. FAO Irrigation and Drainage Paper 56, 15 pp. [Available online at http://www.fao.org/docrep/X0490E/X0490E00.htm.]

  • Cambardella, C. A., T. B. Moorman, J. M. Novak, T. B. Parkin, D. L. Karlen, R. F. Turco, and A. E. Konopka, 1994: Field-scale variability of soil properties in central Iowa soils. Soil. Sci. Soc. Amer. J., 58, 15011511, doi:10.2136/sssaj1994.03615995005800050033x.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Chen, C., C. Lei, A. Deng, C. Qian, W. Hoogmoed, and W. Zhang, 2011: Will higher minimum temperatures increase corn production in northeast China? An analysis of historical data over 1965–2008. Agric. For. Meteor., 151, 15801588, doi:10.1016/j.agrformet.2011.06.013.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Chen, S., Y. Shi, Y. Guo, and Y. X. Zheng, 2010: Temporal and spatial variation of annual mean air temperature in arid and semiarid region in northwest China over a recent 46 year period. J. Arid Land, 2, 8797, doi:10.3724/SP.J.1227.2010.00087.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Ding, Y., and Coauthors, 2007: History and Future Trend of Climate Change: China’s National Assessment Report on Climate Change. Science and Publishing House and Technology Press, 422 pp.

  • Espadafor, M., I. J. Lorite, P. Gavilán, and J. Berengena, 2011: An analysis of the tendency of reference evapotranspiration estimates and other climate variables during the last 45 years in southern Spain. Agric. Water Manage., 98, 10451061, doi:10.1016/j.agwat.2011.01.015.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Estévez, J., P. Gavilán, and J. Berengena, 2009: Sensitivity analysis of a Penman–Monteith type equation to estimate reference evapotranspiration in southern Spain. Hydrol. Processes, 23, 33423353, doi:10.1002/hyp.7439.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Gao, X., C. Wang, J. Zhang, and X. Xue, 2012: Crop water requirement and temporal-spatial variation of drought and flood disaster during growth stages for maize in northeast during past 50 years. Trans. Chin. Soc. Agric. Eng., 28, 101109.

    • Search Google Scholar
    • Export Citation
  • Gong, L., C. Xu, D. Chen, S. Halldin, and Y. D. Chen, 2006: Sensitivity of the Penman–Monteith reference evapotranspiration to key climatic variables in the Changjiang (Yangtze River) basin. J. Hydrol., 329, 620629, doi:10.1016/j.jhydrol.2006.03.027.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Gong, Y., W. S. Lu, and L. Tao, 2003: Spatial and temporal distribution of anomalous temperature/precipitation in spring and summer as well as their relationships to drought/flood in northeast of China. J. Nanjing Inst. Meteor., 26, 349357.

    • Search Google Scholar
    • Export Citation
  • Goyal, R. K., 2004: Sensitivity of evapotranspiration to global warming: A case study of arid zone of Rajasthan (India). Agric. Water Manage., 69, 111, doi:10.1016/j.agwat.2004.03.014.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Guenang, G. M., and F. M. Kamga, 2014: Computation of the standardized precipitation index (SPI) and its use to assess drought occurrences in Cameroon over recent decades. J. Appl. Meteor. Climatol., 53, 23102324, doi:10.1175/JAMC-D-14-0032.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Huang, W. H., and Coauthors, 2009: Analysis of spatio-temporal characteristic on seasonal drought of spring maize based on crop water deficit index. Trans. Chin. Soc. Agric. Eng., 25, 2834.

    • Search Google Scholar
    • Export Citation
  • IPCC, 2007: Summary for policymakers. Climate Change 2007: The Physical Science Basis, S. Solomon et al., Eds., Cambridge University Press, 1–18.

    • Crossref
    • Export Citation
  • IPCC, 2013: Summary for policymakers. Climate Change 2013: The Physical Science Basis, T. F. Stocker et al., Eds., Cambridge University Press, 1–29.

  • Irmak, S., J. O. Payero, D. L. Martin, A. Irmak, and T. A. Howell, 2006: Sensitivity analyses and sensitivity coefficients of standardized daily ASCE-Penman–Monteith equation. J. Irrig. Drain. Eng., 132, 564578, doi:10.1061/(ASCE)0733-9437(2006)132:6(564).

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Kendall, M. G., 1975: Rank Correlation Measures. Charles Griffin, 202 pp.

  • Keyantash, J., and A. J. Dracup, 2002: The quantification of drought: An evaluation of drought indices. Bull. Amer. Meteor. Soc., 83, 11671180, doi:10.1175/1520-0477(2002)083<1191:TQODAE>2.3.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Kim, D., H. Byun, K. Choi, and S. Oh, 2011: A spatiotemporal analysis of historical droughts in Korea. J. Appl. Meteor. Climatol., 50, 18951912, doi:10.1175/2011JAMC2664.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Lee, S., H. Byun, and H. L. Tanaka, 2012: Spatiotemporal characteristics of drought occurrences over Japan. J. Appl. Meteor. Climatol., 51, 10871098, doi:10.1175/JAMC-D-11-0157.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Li, R., A. Tsunekawa, and M. Tsubo, 2014: Index-based assessment of agricultural drought in a semi-arid region of Inner Mongolia, China. J. Arid Land, 6, 315, doi:10.1007/s40333-013-0193-8.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Liang, L., L. Li, L. Zhang, J. Li, and B. Li, 2008: Sensitivity of Penman-Monteith reference crop evapotranspiration in Tao’er River basin of northeastern China. Chin. Geogr. Sci., 18, 340347, doi:10.1007/s11769-008-0340-x.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Liu, Z. J., X. G. Yang, W. F. Wang, K. N. Li, and X. Y. Zhang, 2009: Characteristics of agricultural climate resources in three provinces of northeast China under global climate change. Chin. J. Appl. Ecol., 20, 21992206.

    • Search Google Scholar
    • Export Citation
  • Lv, H., Y. Zhang, J. Wang, S. Zhang, X. Lou, and Y. Zhang, 2009: Classification of agricultural drought category. General Administration of Quality Supervision, Inspection and Quarantine of the People’s Republic of China, Beijing, China, 44 pp.

  • Mann, H. B., 1945: Nonparametric tests against trend. Econometrica, 13, 245259, doi:10.2307/1907187.

  • McCabe, G. J., and D. M. Wolock, 2015: Variability and trends in global drought. Earth Space Sci., 2, 223228, doi:10.1002/2015EA000100.

  • McKenney, M. S., and N. J. Rosenberg, 1993: Sensitivity of some potential evapotranspiration estimation methods to climate change. Agric. For. Meteor., 64, 81110, doi:10.1016/0168-1923(93)90095-Y.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • McVicar, T. R., T. G. Van Niel, L. T. Li, M. L. Roderick, D. P. Rayner, L. Ricciardulli, and R. J. Donohue, 2008: Wind speed climatology and trends for Australia, 1975–2006: Capturing the stilling phenomenon and comparison with near-surface analysis output. Geophys. Res. Lett., 35, 16, doi:10.1029/2008GL035627.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Min, S. K., X. Zhang, F. W. Zwiers, and G. C. Hegerl, 2011: Human contribution to more-intense precipitation extremes. Nature, 470, 378381, doi:10.1038/nature09763.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Mitchell, J. M., B. Dzerdzeevskii, H. Flohn, W. L. Hofmeyr, H. H. Lamb, K. N. Rao, and C. C. Wallen, 1966: Climate change. WMO Tech. Note 79, 79 pp.

  • Monteith, J. L., 1965: Evaporation and the environment. The State and Movement of Water in Living Organism, G. E. Fogg, Ed., Cambridge University Press, 432 pp.

  • Ning, L., and R. S. Bradley, 2015: Snow occurrence changes over the central and eastern United States under future warming scenarios. Sci. Rep., 5, 17073, doi:10.1038/srep17073.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Ning, L., and R. S. Bradley, 2016: NAO and PNA influences on winter temperature and precipitation over the eastern United States in CMIP5 GCMs. Climate Dyn., 46, 12571276, doi:10.1007/s00382-015-2643-9.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Rana, G., and N. Katerji, 1998: A measurement based sensitivity analysis of the Penman–Monteith actual evapotranspiration model for crops of different height and in contrasting water status. Theor. Appl. Climatol., 60, 141149, doi:10.1007/s007040050039.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Rayner, D. P., 2007: Wind run changes: The dominant factor affecting pan evaporation trends in Australia. J. Climate, 20, 33793394, doi:10.1175/JCLI4181.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Roderick, M. L., and G. D. Farquhar, 2002: The cause of decreased pan evaporation over the past 50 years. Science, 298, 14101411, doi:10.1126/science.1075390-a.

    • Search Google Scholar
    • Export Citation
  • Roderick, M. L., M. T. Hobbins, and G. D. Farquhar, 2009a: Pan evaporation trends and the terrestrial water balance. I. Principles and observations. Geogr. Compass, 3, 746760, doi:10.1111/j.1749-8198.2008.00213.x.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Roderick, M. L., M. T. Hobbins, and G. D. Farquhar, 2009b: Pan evaporation trends and the terrestrial water balance. II. Energy balance and interpretation. Geogr. Compass, 3, 761780, doi:10.1111/j.1749-8198.2008.00214.x.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Sheffield, J., and E. F. Wood, 2008: Global trends and variability in soil moisture and drought characteristics, 1950–2000, from observation-driven simulations of the terrestrial hydrologic cycle. J. Climate, 21, 432458, doi:10.1175/2007JCLI1822.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Sonali, P., and D. N. Kumar, 2013: Review of trend detection methods and their application to detect temperature changes in India. J. Hydrol., 476, 212227, doi:10.1016/j.jhydrol.2012.10.034.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Sui, Y., W. H. Huang, X. G. Yang, and M. S. Li, 2012: Characteristics and adaption of seasonal drought in southern China under the background of global climate change. II. Spatiotemporal characteristics of drought for wintering grain- and oil crops based on crop water deficit index. Yingyong Shengtai Xuebao, 23, 24672476.

    • Search Google Scholar
    • Export Citation
  • Sun, F., S. Yang, and P. Chen, 2005: Climatic warming-drying trend in northeastern China during the last 44 years and its effects. Shengtaixue Zazhi, 24, 751755.

    • Search Google Scholar
    • Export Citation
  • Sun, F., J. Yuan, and S. Lu, 2006: The change and test of climate in northeast China over last 100 years. Climate Environ. Res., 11, 101107.

    • Search Google Scholar
    • Export Citation
  • Sun, F., S. Yang, and G. Ren, 2007: Decade variations of precipitation event frequency, intensity and duration in the northeast China. J. Appl. Meteor. Sci., 18, 610618.

    • Search Google Scholar
    • Export Citation
  • Tabari, H., and P. H. Talaee, 2014: Sensitivity of evapotranspiration to climatic change in different climates. Global Planet. Change, 115, 1623, doi:10.1016/j.gloplacha.2014.01.006.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Tong, L., S. Kang, L. Zhang, and L. Zhang, 2007: Temporal and spatial variations of evapotranspiration for spring wheat in the Shiyang River basin in northwest China. Agric. Water Manage., 87, 241250, doi:10.1016/j.agwat.2006.07.013.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wang, H., and Coauthors, 2012: Extreme climate in China: Facts, simulation and projection. Meteor. Z., 21, 279304, doi:10.1127/0941-2948/2012/0330.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wen, K., and Coauthors, 2005: China’s Meteorological Disaster Encyclopedia. Meteorological Press, 322 pp.

  • Xie, W. J., X. G. Yang, J. Yang, L. M. Liu, Q. Ye, C. Y. Dong, Z. J. Liu, and J. Zhao, 2014: Spatio-temporal characteristics of drought for soybean under climate change in the three provinces of northeast China. Acta Ecol. Sin., 34, 62326243.

    • Search Google Scholar
    • Export Citation
  • Xu, C., L. Gong, T. Jiang, D. Chen, and V. P. Singh, 2006: Analysis of spatial distribution and temporal trend of reference evapotranspiration and pan evaporation in Changjiang (Yangtze River) catchment. J. Hydrol., 327, 8193, doi:10.1016/j.jhydrol.2005.11.029.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Xue, Q., 2013: Analysis on the change of 30 year’s soybean areas, production and yield in China and northeast China. Zhongguo Nongxue Tongbao, 29, 102106.

    • Search Google Scholar
    • Export Citation
  • Yan, C. J., W. B. Wang, X. J. Tu, C. L. Wang, L. J. Zhang, Q. Du, and S. H. Song, 2013: Effect of drought stress at different growth stage on yield and root characteristics of soybean. Dadou Kexue, 32, 5962.

    • Search Google Scholar
    • Export Citation
  • Yang, X., L. Ren, Y. Liu, D. Jiao, and S. Jiang, 2014: Hydrological response to land use and land cover changes in a sub-watershed of west Liaohe River basin, China. J. Arid Land, 6, 678689, doi:10.1007/s40333-014-0026-4.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Yin, X. G., J. E. Olesen, M. Wang, I. Öztürk, and F. Chen, 2016: Climate effects on crop yields in the northeast farming region of China during 1961–2010. J. Agric. Sci., 154, 11901208, doi:10.1017/S0021859616000149.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Zhang, M., J. He, B. Wang, S. Wang, S. Li, W. Liu, and X. Ma, 2013: Extreme drought changes in southwest China from 1960 to 2009. J. Geogr. Sci., 23, 316, doi:10.1007/s11442-013-0989-7.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Zhang, S., Y. Zhang, R. Ji, F. Cai, and J. Wu, 2011: Analysis of spatio-temporal characteristics of drought for maize in northeast China. Agric. Res. Arid Areas, 29, 231236.

    • Search Google Scholar
    • Export Citation
  • Zhang, X., S. Kang, L. Zhang, and J. Liu, 2010: Spatial variation of climatology monthly crop reference evapotranspiration and sensitivity coefficients in Shiyang River basin of northwest China. Agric. Water Manage., 97, 15061516, doi:10.1016/j.agwat.2010.05.004.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Zhang, Y., H. Lu, and S. Li, 2008: Applicability of crop water deficit index in agricultural drought monitoring. Mater. Sci. Technol., 36, 596600.

    • Search Google Scholar
    • Export Citation
  • Zhao, C., G. Ren, and Y. Zhang, 2009: Climate change of the northeast China over the past 50 years. Ganhanqu Ziyuan Yu Huanjing, 23, 2530.

    • Search Google Scholar
    • Export Citation
  • View in gallery

    Location of northeast China and the meteorological stations used in the study.

  • View in gallery

    The CWDI, ETc, and precipitation (labeled Pr) for soybeans during the growth season averaged from 1981 to 2010 in northeast China.

  • View in gallery

    Interannual variations of the CWDI for soybeans during the seedling–branching, flowering, podding, and filling stages in northeast China (1981–2010).

  • View in gallery

    Five types of soybean-related drought frequency at each growth stage in northeast China.

  • View in gallery

    Spatial distribution of soybean-related drought frequency during the (a) seedling–branching, (b) flowering, (c) podding, and (d) filling stages in northeast China.

  • View in gallery

    Percent change in the CWDI with respect to percent changes in the meteorological variables for the average of all stations in northeast China.

  • View in gallery

    Spatial distribution of the SC for relative humidity, air temperature, precipitation, wind speed, and sunshine hours.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 311 223 2
PDF Downloads 132 66 1

Spatiotemporal Assessment of Drought Related to Soybean Production and Sensitivity Analysis in Northeast China

View More View Less
  • 1 Institute of Environment and Sustainable Development in Agriculture, Chinese Academy of Agricultural Science/National Engineering Laboratory of Efficient Water Use, Disaster Resistant and Mitigation of Crops/Key Laboratory of Agricultural Environment, Ministry of Agriculture, Beijing, China
Full access

Abstract

Drought is a typical disaster in the main soybean production area of northeast China. The spatiotemporal variations of drought related to soybean production based on a crop water deficit index (CWDI) and sensitivity to meteorological variables were investigated in northeast China using daily meteorological data from 87 weather stations from 1981 to 2010. Statistical analysis revealed that precipitation could not meet the water demands of soybeans during the seedling–branching, filling, and maturing stages, and excessive drought occurred more often in northeast China. The Mann–Kendall test indicated that the soybean CWDI significantly increased during the filling stage. Kriging spatial analysis showed that the most drought-prone area was located in the west of northeast China. Explanations for the spatiotemporal variations of the drought for soybean production were explored in terms of meteorological variables. Statistical analysis showed that the crop evapotranspiration, air temperature, wind speed, and number of sunshine hours were significantly higher and the precipitation and relative humidity were significantly lower in the drought-prone area than in the dry area less prone to droughts. An explored method of sensitive analysis quantitatively revealed that precipitation and humidity negatively affected the CWDI, whereas temperature, wind speed, and number of sunshine hours positively affected the CWDI. The CWDI was most sensitive to precipitation. These results not only provide valuable information for soybean planning and management but also produce important background and physical evidence for the influence of climate on the drought related to soybean production in northeast China.

Denotes content that is immediately available upon publication as open access.

© 2017 American Meteorological Society. For information regarding reuse of this content and general copyright information, consult the AMS Copyright Policy (www.ametsoc.org/PUBSReuseLicenses).

Corresponding author e-mail: BuChun Liu, liubuchun@caas.cn

Abstract

Drought is a typical disaster in the main soybean production area of northeast China. The spatiotemporal variations of drought related to soybean production based on a crop water deficit index (CWDI) and sensitivity to meteorological variables were investigated in northeast China using daily meteorological data from 87 weather stations from 1981 to 2010. Statistical analysis revealed that precipitation could not meet the water demands of soybeans during the seedling–branching, filling, and maturing stages, and excessive drought occurred more often in northeast China. The Mann–Kendall test indicated that the soybean CWDI significantly increased during the filling stage. Kriging spatial analysis showed that the most drought-prone area was located in the west of northeast China. Explanations for the spatiotemporal variations of the drought for soybean production were explored in terms of meteorological variables. Statistical analysis showed that the crop evapotranspiration, air temperature, wind speed, and number of sunshine hours were significantly higher and the precipitation and relative humidity were significantly lower in the drought-prone area than in the dry area less prone to droughts. An explored method of sensitive analysis quantitatively revealed that precipitation and humidity negatively affected the CWDI, whereas temperature, wind speed, and number of sunshine hours positively affected the CWDI. The CWDI was most sensitive to precipitation. These results not only provide valuable information for soybean planning and management but also produce important background and physical evidence for the influence of climate on the drought related to soybean production in northeast China.

Denotes content that is immediately available upon publication as open access.

© 2017 American Meteorological Society. For information regarding reuse of this content and general copyright information, consult the AMS Copyright Policy (www.ametsoc.org/PUBSReuseLicenses).

Corresponding author e-mail: BuChun Liu, liubuchun@caas.cn

1. Introduction

According to the Fifth Assessment Report by the Intergovernmental Panel on Climate Change (IPCC), the global surface air temperature increased linearly by 0.65° ± 1.06°C over 133 yr (1880–2012). Most of this observed temperature increase has occurred since the middle of the twentieth century, primarily because the greenhouse gas concentrations in the atmosphere increased because of human activities, for example, the burning of fossil fuels and deforestation (IPCC 2013). This temperature increase is likely to continue in the twenty-first century, resulting in a total temperature increase of 1.1°–6.4°C, based on different emission scenarios (IPCC 2007). In China, the average recorded temperature increased by 0.5°–0.8°C during the last century, and a temperature increase of 1.1°C occurred in the last 50 yr (Ding et al. 2007; Chen et al. 2010). Climate change has caused an increased occurrence of drought, flood, frost, and other meteorological disasters worldwide (Sheffield and Wood 2008; Min et al. 2011; Wang et al. 2012; IPCC 2013; Sonali and Kumar 2013; Zhang et al. 2013; Ning and Bradley 2015; Ning and Bradley 2016).

Agriculture is directly affected by meteorological disasters. The mainly rain-fed northeast China is the largest soybean production area in China, with the soybean planting area and yield in this region accounting for 40%–50% of the total soybean production in China (Xue 2013). The soybean production in this area suffers from drought seriously (Sun et al. 2007; Zhang et al. 2011). Precipitation in northeast China has exhibited an uneven and extreme pattern, which led to the expectation of increased drought (Gong et al. 2003; Sun et al. 2005, 2007; Zhao et al. 2009; Yang et al. 2014), although no change in global drought for 1901 through 2009 was reported by McCabe and Wolock (2015).

Numerous drought indices have been used to evaluate droughts, such as the Palmer drought severity index, standardized precipitation index, relative soil moisture, effective drought index, and accumulated crop water deficit index (Kim et al. 2011; Lee et al. 2012; Guenang and Kamga 2014; Li et al. 2014). These indices can be clustered into three categories: those measuring meteorological, hydrological, or agricultural droughts (Keyantash and Dracup 2002). Most studies were not focused on agriculture, for which meteorological drought indices are commonly used to evaluate the spatiotemporal characteristics of drought (Kim et al. 2011; Lee et al. 2012; Guenang and Kamga 2014). The accumulated crop water deficit index, one of the agricultural drought indexes, has been used in assessing the spatiotemporal variation of crop drought in China (Zhang et al. 2008; Huang et al. 2009; Zhang et al. 2011; Gao et al. 2012; Sui et al. 2012; Xie et al. 2014). Zhang found that maize drought had increased significantly in northeast China and occurred in each stage of maize development (Zhang et al. 2011). Huang showed that drought for spring maize in Hunan Province in China occurred with the highest frequency in the midsouth part of Hunan, followed by the east and north parts, and there was a low drought frequency region located in the west part (Huang et al. 2009). Sui found that drought of wintering crops occurred more frequently in southwest China, north Huaihe basin, and parts of south China during the growing season (Sui et al. 2012). The accumulated crop water deficit index is defined as the sum of five consecutive 10-day periods of weighted crop water deficit index; the first four 10-day periods’ accumulated crop water deficit index of the crop growing season usually cannot be calculated unless the crop coefficient from the four 10-day periods before the growing season was assumed, leading to decreased integrity and accuracy of the result. Therefore, the non-accumulated crop water deficit index (CWDI) was chosen in the present study that reflects the combined effects of plant and climate factors on crop water stress.

A sensitivity analysis is required to understand the relative importance of meteorological variables to soybean-related drought. Results from sensitivity analyses are of great importance to determine the approximate change in soybean-related drought expected for a known change in one of the meteorological variables and are also necessary to determine the most important meteorological variable that affects CWDI (Irmak et al. 2006). The sensitivity of reference evapotranspiration (ET0) to meteorological factor analysis in different climates has been calculated for a single station or spatially by several scientists (Rana and Katerji 1998; Goyal 2004; Irmak et al. 2006). Gong et al. (2006) conducted a sensitivity analysis of ET0 to key meteorological variables in Changjiang basin and derived the spatial variation of sensitivity coefficients by interpolating the station estimates. They found that relative humidity was the most sensitive variable, followed by shortwave radiation, air temperature, and wind speed, and the sensitivity coefficients of wind speed and shortwave radiation had opposite spatial patterns. Zhang et al. (2010) mapped the sensitivity coefficients of ET0 to the climatic variables of selected months in the Shiyang River basin of northwest China. They found that each climatic variable fluctuates during the year and varied in the studied region. Tabari and Talaee (2014) inferred that the sensitivity of ET0 to the same climatic variables significantly differed among humid, cold semiarid, warm semiarid, and arid climates, and the sensitivity of ET0 to wind speed and air temperature decreased from arid to humid climates, whereas its sensitivity to sunshine hours increased from arid to humid environments.

The objectives of this study were to investigate the spatiotemporal variation of soybean-related drought based on CWDI and estimate the sensitivity of soybean-related drought to meteorological variables in northeast China. This information not only provides valuable reference for soybean planning and management but also produces important background and physical evidence for the influence of climate on soybean-related drought in northeast China.

2. Study area and data analysis

a. Study area

The study area is located in northeast China (118°53′–135°05′E and 38°43′–53°33′N) and includes Heilongjiang, Jilin, and Liaoning Provinces, with a total area of 788 000 km2 (Fig. 1). The annual precipitation is 400–1000 mm, and 80% of the precipitation falls between May and September. The mean annual air temperature ranges from −4.3° to 10.8°C, and the frost period typically begins on 2 October and ends on 28 March of the following year (Chen et al. 2011). The annual total sunshine hours are in the range of 2212–2955 h (Liu et al. 2009).

Fig. 1.
Fig. 1.

Location of northeast China and the meteorological stations used in the study.

Citation: Journal of Applied Meteorology and Climatology 56, 4; 10.1175/JAMC-D-16-0195.1

b. Climate and crop data

Daily meteorological data during the period 1981–2010, including relative humidity, mean air temperature, precipitation, wind speed, and sunshine hour data, were collected from 87 weather stations in northeast China. In total, 25 daily data for relative humidity, 27 for mean air temperature, 67 for wind speed, and 690 for the number of sunshine hours were missing. The missing data were replaced with the daily average value from the 1981–2010 period. These data were used to calculate the CWDI, crop evapotranspiration (ETc) characteristics of soybean-related drought, and meteorological factors sensitivity to CWDI by International Business Machines Corporation’s (IBM) SPSS Statistics 17.0 software and Microsoft Corporation’s Microsoft Excel 2010 software. ETc was used to approximate soybean water demand (Allen et al. 1998), and precipitation was considered to be equal to water supply because soybean production in northeast China is mainly rain fed.

Records of crop phenology (seedling, three leaves, branching, flowering, podding, filling, and maturing dates) were collected from 10 selected agrometeorological observation stations in northeast China during the period 1991–2011. The averaged dates of seedling, three leaves, branching, flowering, podding, filling, and maturing were used for the entire period from 1981 to 2010 in northeast China. The seedling stage to the filling stage was chosen because drought can affect soybean production during this period. However, the study of six growth stages was slightly complicated. For the convenience of analysis, the seedling, three leaves, and branching growth stages were collapsed because they are not critical growth stages for water requirement relative to the flowering, podding, and filling stages (Lv et al. 2009; Table 1).

Table 1.

Crop coefficient and growth stages of soybean. Early indicates from the first day to the 10th day of a month, middle indicates from the 11th day to the 20th day of a month, and late indicates from the 21st day to the end of the month.

Table 1.

c. CWDI calculation

This study quantifies drought severity using the CWDI. The soybean CWDI was defined as the percentage of the difference between water supply and demand divided by crop water demand, as follows:
e1
where Pj is the amount of precipitation that occurred during a given period (mm), which approximates the crop water supply, and ETcj is the crop evapotranspiration of a given period (mm), which approximates the crop water demand.
ETc was determined using the crop coefficient-reference evapotranspiration procedure. The reference evapotranspiration (ET0) was multiplied by the crop coefficient (kc) to estimate crop evapotranspiration as follows:
e2

The crop coefficients in this study refer to the results of the classification of agricultural drought categories (Lv et al. 2009), as shown in Table 1.

ET0 was estimated using the FAO-56 Penman–Monteith equation (Allen et al. 1998), a simplification of the original Penman–Monteith equation (Monteith 1965), and was expressed as follows:
e3
where ET0 is the reference evapotranspiration (mm day−1), Δ is the vapor pressure slope curve (kPa °C−1), Rn is the net radiation at the crop surface (MJ m−2 day−1), and G is the soil heat flux density (MJ m−2 day−1), which was assumed to be zero based on the FAO-56 recommendation that the magnitude of the day or 10-day soil heat flux beneath the grass reference surface may be ignored because it is relatively small. The quantity T is the mean daily air temperature at a height of 2 m (°C), U2 is the wind speed at 2 m above the ground (m s−1), ed is the saturation vapor pressure (kPa), ea is the actual vapor pressure (kPa), and γ is the psychrometric constant (kPa °C−1).

d. Drought duration

Drought duration is the time between the onset and the end date. The drought onset date was determined to be the day on which the CWDI > 15 in the seedling–branching stage or CWDI > 10% in the flowering, podding, or filling stage (critical growth stage of water requirement), and the end date was determined to be the day prior to that on which the CWDI ≤ 15% in the seedling–branching stage or CWDI ≤ 10% in the flowering, podding, or filling stage.

e. Frequency calculation of classified droughts

The soybean CWDI of each growth stage for the 87 meteorological stations during the period of 1981–2010 was calculated to understand the frequency of soybean-related drought occurrence and the spatial distribution of drought in northeast China. The classification scale of the CWDI was showed in Table 2.

Table 2.

Classifications of soybean-related drought by the CWDI.

Table 2.

The frequency of each drought type was computed using the following equation according to the threshold values shown in Table 2:
e4
where F is the drought frequency of a given growth stage, N is the number of times a drought occurred during a given growth stage based on the drought index classification, and n is the number of years investigated in a given growth stage.

f. Trend test

The nonparametric Mann–Kendall test (Mann 1945; Kendall 1975) is one of the most widely used nonparametric tests in climate studies for detecting trends in time series (Sheffield and Wood 2008; Espadafor et al. 2011; Zhang et al. 2011; Sonali and Kumar 2013) and is highly recommended for general use by the World Meteorological Organization (Mitchell et al. 1966). The test was used to detect increasing or decreasing trends in the CWDI data series and meteorological parameters by MathWorks’s MATLAB 7.0 software. The theoretical framework of the Mann–Kendall test statistic is as follows:
e5
where xj and xz are two generic sequential data values of the variable in question, N is the length of the dataset, and sign(y) takes the following values:
e6
Assuming a normal distribution, the variance is given by the following relation:
e7
where ti is the number of occurrences of extent i.
Therefore, a normalized test statistic Z could be computed as follows:
e8
A positive value of Z indicates an increasing trend, and a negative value of Z indicates a decreasing trend, while a zero value of Z indicates no trend. At the 1% significance level, the null hypothesis of no trend is rejected if |Z| > 2.576; at the 5% significance level, the null hypothesis of no trend is rejected if |Z| > 1.96; at the 10% significance level, the null hypothesis of no trend is rejected if |Z| > 1.645.

g. Sensitivity analysis

Sensitivity analysis was used to analyze the relative changes of ET0 against changes of the meteorological variable (McKenney and Rosenberg 1993; Goyal 2004; Irmak et al. 2006; Zhang et al. 2010). To explore the sensitivity of the CWDI to meteorological variables, we modified the sensitivity coefficient equation as follows:
e9
where SC is the sensitivity coefficient, ΔCWDI is the unit change in the CWDI, ΔCV is the unit change in the meteorological variable CWDI, and CV is the base value before change. The SC could represent the changes in CWDI caused by the meteorological variable if the sensitivity curve is linear (Gong et al. 2006; Zhang et al. 2010). Sensitivity analyses of the CWDIs in each growth stage were conducted at each station from −20% to +20% at an interval of 5% (eight scenarios) for each of the five variables while holding all other parameters constant, and the sensitivity curve was plotted by the average ΔCWDI/CWDI of all stations against the variations of the meteorological variable.

h. Spatial interpolation

To know the general distribution characters of soybean-related drought and sensitivity of meteorological variable in northeast China, ordinary kriging with sphere variogram was selected for interpolation of the soybean cumulative drought frequency and the SC (meteorological variable sensitivity to CWDI) of the 87 stations into a grid of 1 km × 1 km in latitude and longitude by Environmental Systems Research Institute’s (ESRI) Arc-Map 10.0 geostatistical analysis software. There are three parameters for the spherical variogram models: nugget (C0), sill (C0 + C1), and range. The nugget-to-sill ratio (%) reflects the extent of spatial autocorrelation, with a ratio of <25% indicating strong spatial dependence, a ratio of 25%–75% indicating moderate spatial dependence, and a ratio of >75% indicating weak spatial dependence (Cambardella et al. 1994). In the present study, the nugget-to-sill ratios are less than 75% indicating that the interpolated factors have strong–moderate spatial structure. The ranges are greater than the weather-station interval, indicating that the samples can satisfy this study.

3. Results

a. Soybean water demand and supply during the growing season in northeast China

Precipitation, ETc, and the CWDI during the soybean growing season were averaged over northeast China for the period of 1981–2010 and are shown in Fig. 2. Average precipitation increased from 18 mm at the beginning of the seedling stage to 64 mm at the beginning of the podding stage and then gradually decreased to 13 mm at the end of the growing season. The soybean ETc exhibited a pattern similar to that of precipitation. The CWDI of soybean decreased from the start of the seedling–branching stage and then remained less than zero during the flowering and podding stages before the CWDI increased during the filling stage and maturing stage.

Fig. 2.
Fig. 2.

The CWDI, ETc, and precipitation (labeled Pr) for soybeans during the growth season averaged from 1981 to 2010 in northeast China.

Citation: Journal of Applied Meteorology and Climatology 56, 4; 10.1175/JAMC-D-16-0195.1

b. Characteristics of soybean-related drought in each growth stage

To further analyze soybean-related drought, the annual CWDI, drought duration, and frequency of different drought types were analyzed in each growth stage. Annual CWDI and drought duration are shown in Fig. 3 and Table 3, respectively. The annual CWDI in the filling stage was higher than those during the seedling–branching, flowering, and podding stage (Fig. 3). In most cases, the higher annual CWDI was well matched with the longer drought duration in each growth stage (Fig. 3 and Table 3). For the seedling–branching stage, the annual CWDI was higher in northeast China. In 2000, the annual CWDI was 60%, and, in 1982, the annual CWDI was 52% in northeast China (Fig. 3). The longest droughts occurred in 2007 and lasted for 23 days. The drought durations in 1982 and 2000 were 19 and 15 days, respectively; they were the fourth and third longest droughts (Table 3). For the flowering stage and podding stage, the annual CWDI were normally lower than that of the seedling–branching stage because these two stages occur in the rainy season of northeast China. However, soybean-related drought still occurred sometimes. In these two stages, the highest annual CWDI and the longest drought occurred in the same years in northeast China, which were 43% and 14 days in the flowering stage in 1992 and 30% and 23 days in the podding stage in 2009 (Fig. 3 and Table 3). For the filling stage, the highest annual CWDI were 66% in 2001. Drought lasting for more than 10 days occurred two times in 2001, although the longest drought did not occur in that year.

Fig. 3.
Fig. 3.

Interannual variations of the CWDI for soybeans during the seedling–branching, flowering, podding, and filling stages in northeast China (1981–2010).

Citation: Journal of Applied Meteorology and Climatology 56, 4; 10.1175/JAMC-D-16-0195.1

Table 3.

Top 10 soybean-related drought durations (D, days) in each growth stage in northeast China.

Table 3.

The frequencies of light, medium, and severe drought in the flowering and podding stages were significantly lower than those during the seedling–branching stage and filling stage (Fig. 4). Excessive drought occurred more often in the filling stage than in the other growth stages. Regardless of the drought type, the soybean-related drought frequency was between 36% and 61% from the seedling–branching to filling states in northeast China, and the frequency of drought in the filling stage was significantly higher than in the other stages.

Fig. 4.
Fig. 4.

Five types of soybean-related drought frequency at each growth stage in northeast China.

Citation: Journal of Applied Meteorology and Climatology 56, 4; 10.1175/JAMC-D-16-0195.1

c. Temporal variation of the CWDI and related meteorological variables in each soybean growth stage

The Mann–Kendall analysis showed that the CWDI in the seedling–branching, flowering, and podding stages have exhibited insignificantly increasing trends in northeast China since the 1980s (Table 4 and Fig. 3). In contrast to the three soybean pre-growth stages, the CWDI trended upward significantly in the filling stage (Table 4 and Fig. 3).The same Mann–Kendall test analysis procedure was performed on the related meteorological variables. The temperature showed significant increasing trends and wind speed showed significant decreasing trends in each growth stage (Table 4). In the filling stage, the related meteorological variables showed different trends. The ETc and air temperature have significant increasing trends; precipitation, relative humidity, and wind speed have significant decreasing trends.

Table 4.

Linear trend of the CWDI, relative humidity, air temperature, precipitation, wind speed and sunshine hours at each soybean growth stage in northeast China. Asterisks represent the significance of the trend according to the Mann–Kendall test: *, **, and *** indicate significance levels of 0.10, 0.05, and 0.01, respectively. Positive values in the table correspond to upward trends, and negative values correspond to downward trends.

Table 4.

d. Spatial distribution of soybean-related drought in northeast China

The spatial distributions of the frequency of soybean-related drought during each growth stage are shown in Fig. 5. For the seedling–branching stage, the average drought frequency in northeast China was 51% (Fig. 4). Areas with drought frequencies higher than 50% but lower than 75% were located in the central and western parts of northeast China and east of Heilongjiang, in the remaining areas, the drought frequencies were between 25% and 50% (Fig. 5a). For the flowering stage, the drought frequency was 39% in northeast China (Fig. 4), which was lower than the drought frequency in the seedling–branching stage. The areas with the highest drought frequencies were located in the western parts of northeast China and northeast of Heilongjiang, and the lowest drought frequencies occurred in the eastern parts of Jilin and Liaoning (Fig. 5b). For the podding stage, the drought frequency was 37% in northeast China (Fig. 4), which was lower than the drought frequency during the other growth stages. The area with high drought frequency was located southwest of Heilongjiang, and the southeastern regions of northeast China had the lowest drought frequency (Fig. 5c). For the filling stage, the drought frequency was 60% in northeast China (Fig. 4), which was significantly higher than the drought frequency observed for the other growth stages. The high drought frequency areas were located in southwestern Heilongjiang and western Liaoning (Fig. 5d). The soybean-related drought frequency in most of northeast China was higher than 50%, and only the southeastern and northeastern regions of northeast China had soybean-related drought frequencies under 50%.

Fig. 5.
Fig. 5.

Spatial distribution of soybean-related drought frequency during the (a) seedling–branching, (b) flowering, (c) podding, and (d) filling stages in northeast China.

Citation: Journal of Applied Meteorology and Climatology 56, 4; 10.1175/JAMC-D-16-0195.1

The related meteorological variable was analyzed based on the spatial distributions of the frequency of soybean-related drought (Table 5). The high drought frequency areas during each growth stage had low relative humidity and precipitation and high ETc, air temperature, wind speed, and number of sunshine hours. In contrast, the low drought frequency areas had a high relative humidity and precipitation and low ETc, air temperature, wind speed and number of sunshine hours, and areas with medium drought frequency experienced medium ETc, relative humidity, air temperature, precipitation, wind speed, and number of sunshine hours. These values of same meteorological variables in same growth stage in different drought frequency areas show significant differences at the 5% significance level.

Table 5.

Averages of meteorological factors at each soybean growth stage in different drought frequency (F) areas. The same meteorological variables in the same soybean growth stage in different drought frequency areas were compared; a, b, and c indicate the difference on the significance level of 0.05 according to ANOVA analysis (same letter for two entries means no significant difference; different letters mean significant difference).

Table 5.

e. Sensitivity of the soybean crop water deficit index to meteorological variables

To further investigate the influences of meteorological variables on soybean-related drought, a sensitivity analysis of CWDI to meteorological variables was explored based on previous ET0 sensitivity analysis (McKenney and Rosenberg 1993; Goyal 2004; Irmak et al. 2006; Zhang et al. 2010). This method can quantitatively detect the effects of each meteorological variable on soybean-related drought. The sensitivity curves between the percent changes in each meteorological variable and the relative percent changes in the CWDI were generally linear (Fig. 6); therefore, the SC of the meteorological variable was calculated as approximately the average value of the eight scenarios according to section 2g. Each meteorological variable fluctuates during the soybean growth period. The slopes of the meteorological variables indicate that the precipitation and relative humidity negatively affected CWDI, while air temperature, wind speed, and sunshine hours positively affected CWDI. Precipitation was the most sensitive variable that influenced the CWDI, and wind speed was the least sensitive variable.

Fig. 6.
Fig. 6.

Percent change in the CWDI with respect to percent changes in the meteorological variables for the average of all stations in northeast China.

Citation: Journal of Applied Meteorology and Climatology 56, 4; 10.1175/JAMC-D-16-0195.1

To detect the effects of each meteorological variable on soybean-related drought quantitatively, the mean SC of each meteorological variable and significant differences among each growth stage were analyzed. The average SC of precipitation, relative humidity, air temperature, and sunshine hours and wind speed in northeast China during the soybean growing season were −4.09, −0.62, 0.60, 0.33, and 0.08, respectively. These meteorological variables showed significant difference at the 5% significance level between each other in addition to the relative humidity and air temperature. The SC of the meteorological variables varied with soybean growth stage (Table 6). The SC of the air temperature and number of sunshine hours during the flowering and podding stages was significantly higher than that during the seedling–branching and filling stages, and the opposite situation was observed for the SC of wind speed. This result indicated that air temperature and number of sunshine hours had greater effects on the CWDI during the middle two soybean growth stages and that wind played an important role on the CWDI during the first and last growth stages. The absolute SC of precipitation was higher during the previous three stages than during the last stage, and the absolute SC of relative humidity was significantly higher during the last three stages than during the first stages. Thus, the CWDI was more sensitive to precipitation before filling and to humidity after seedling–branching.

Table 6.

Sensitivity coefficients and standard deviation for relative humidity, air temperature, wind speed, and sunshine hours in each soybean growth stage. The same meteorological variables in different soybean growth stages were compared; a, b, and c indicate the difference on the significance level of 0.05 according to ANOVA analysis (same letters or including same letters for two entries means no significant difference; different letters mean significant difference).

Table 6.

The spatial distributions of SC are presented in Fig. 7. The SC was different within locations in each growth stage, and the spatial distribution of the SC also changed with soybean growth stage. During the soybean growth stages, the SC of air temperature and wind speed shows a beltlike distribution from southeast to northwest with higher SC in the west area. The SC of relative humidity shows a similar distribution pattern, with lower SC in the southeast area. The SC of precipitation shows a beltlike distribution from east to west in the seedling–branch and filling stages, with higher SC in the east, and a beltlike distribution from south to north in the flowering and podding stages, with lower SC located in the southeast. The SC of sunshine hours shows different distribution pattern during the growing season. During the seedling–branching stage, the SC of sunshine hours shows a beltlike distribution from southeast to northwest in the southern area and a beltlike distribution from southwest to northeast in the northern area, and with higher values located in the southeast and north area; during the flowering and filling stages, SC shows a beltlike distribution from south to north with higher SC in the north and south respectively; during the podding stage, SC shows a beltlike distribution from east to west with higher SC in the west.

Fig. 7.
Fig. 7.

Spatial distribution of the SC for relative humidity, air temperature, precipitation, wind speed, and sunshine hours.

Citation: Journal of Applied Meteorology and Climatology 56, 4; 10.1175/JAMC-D-16-0195.1

4. Discussion

The present study shows that the supply of precipitation could not meet the soybean water demand during the seedling–branching, filling, and maturing stages. The main reason is that the rainy season in northeast China occurred between July and August. The CWDI remained less than zero from early July to the middle of August because the amount of precipitation increased, and the CWDI was greater than zero during other portions of the growing season mostly because of precipitation shortages when compared with soybean evapotranspiration. Previous studies revealed that drought occurring during the filling period could result in large yield losses (Yan et al. 2013). Therefore, the drought with a high CWDI that occurred from late August to the middle of September in the present study would have severe effects on soybean yield.

To understand the soybean-related drought deeply, the characteristics (frequency, severity, and duration) of soybean-related drought in each growth stage were explored. Among these types of drought, excessive drought occurred more often than light, medium, and severe drought during the four growth stages. This pattern indicated that soybean-related drought was very serious in northeast China. Thus, a concerted effort must be made to address drought in these areas. The analysis of annual CWDI and drought duration of soybean shows that these research findings matched the reality of drought that occurred in northeast China well. According to the records of China’s Meteorological Disaster Encyclopedia (Wen et al. 2005), annual precipitation decreased sharply in 1982, which corresponded to a drought of rare severity (when compared with the last 300 years) in Heilongjiang. In 2000, northeast China suffered a rare drought from May to July; Zhang et al. (2011) showed that this drought caused grain production to decrease by 77% relative to the average of the previous five years. In 2001, the south-central part of northeast China experienced drought. In 2007, the drought area was increased and soybean production was decreased in northeast China according the disaster database established by China’s Ministry of Agriculture (http://202.127.42.157/moazzys/zaiqing.aspx). In the 1990s, the south-central region of northeast China experienced a prolonged and severe drought, which had previously rarely occurred (Sun et al. 2007).

In addition to the characteristic of soybean-related drought analysis, trends of CWDI and related meteorological and spatial variation of soybean-related drought were also analyzed. Previous studies showed that relative humidity had a negative effect on the ET0, which is one of the parameters of CWDI, and air temperature, wind speed, and sunshine hours had a positive effect (Xu et al. 2006; Liang et al. 2008). The air temperature increased and wind speed decreased significantly in the four soybean growth stages, in line with other studies showing northeast China has experienced an estimated temperature increase of 1.43°C, which is approximately twice the global increase and double to triple the overall increase in China during the last century (Sun et al. 2006), and the global wind speed had a downward trend over the last 30–50 yr for a range of midlatitude regions (Roderick and Farquhar 2002; Xu et al. 2006; Rayner 2007; McVicar et al. 2008; Roderick et al. 2009a,b). In the present study, precipitation exhibited a decreasing trend in each growth stage but only significantly in the filling stage. Other studies also showed that the annual precipitation and the precipitation that occurred during the crop growing season were expected to decrease in northeast China (Gong et al. 2003; Sun et al. 2005, 2007; Zhao et al. 2009; Yang et al. 2014). By examining the changes of climatic factors during the crop growing season in northeastern China, Yin et al. (2016) came to the conclusion that precipitation is increased insignificantly before soybean flowering and decreased insignificantly during and after flowering during 1961–2010. However, in present study, precipitation decreased insignificantly before soybean flowering, and this difference might be due to the different study period, which was from 1981 to 2010. In the other growth stages, the downward precipitation trend was similar to that of Yin et al. (2016); particularly, in the filling stage, the precipitation showed a significantly decreasing trend from 1981 to 2010.

The spatial analysis of soybean-related drought shows that the soybean-related drought varied spatially in each growth stage. To investigate reasons for the drought distribution, related meteorological variables were analyzed. Combining the meteorological variables with the spatial distribution maps of drought produces important physical evidence for the influences of climate on soybean-related drought in this region. The meteorological variables were quite compatible with the drought distribution in each soybean growth stage. The ETc, air temperature, wind speed, and number of sunshine hours were higher and the precipitation and relative humidity were lower in the dry area than in the less dry areas. Thus, increased ETc, air temperature, wind speed, and number of sunshine hours and decreased relative humidity and precipitation were responsible for the higher drought frequency area, whereas decreased ETc, air temperature, wind speed, and number of sunshine hours and increased relative humidity and precipitation were responsible for the lower drought frequency area. The effectiveness of these meteorological factors on drought was similar to the effectiveness of the climatological factors on ET0 (Xu et al. 2006).

To further understand the influences of meteorological variables on soybean-related drought, the quantitative sensitivity of soybean-related drought to related meteorological variables was investigated. The results show that CWDI was most sensitive to precipitation, followed by relative humidity, air temperature, sunshine hours, and wind speed in northeast China. The CWDI calculation is a three-step approach that quantifies the soybean water shortage through the calculation of the ET0, the contribution of surface characteristics through a crop coefficient, and the water supply from precipitation in the rain-fed area. A meteorological variable that is sensitive to ET0 is sensitive to CWDI as well. Relative humidity, air temperature, sunshine hours, and wind speed have different effects on ET0 estimations. Temperature affects the vapor pressure deficit, net longwave radiation, and slope of the saturation vapor pressure. Relative humidity affects actual vapor pressure and, therefore, the vapor pressure deficit and net longwave radiation. Solar radiation only affects net radiation estimations, and wind speed only affects the aerodynamic term. Precipitation is directly an input variable and only affects CWDI. The magnitude of SC values in the present study was similar to that in the study of Estévez et al. (2009), who conducted a sensitivity analysis to predict responses of ET0 estimated by the FAO-56 Penman–Monteith equation to perturbations of temperature, relative humidity, and wind speed in southern Spain. Gong et al. (2006) used the same approach in the Changjiang (Yangtze River) basin in China and also found low sensitivity coefficients for wind speed throughout the year. However, it is difficult to compare these results with those reported in other works because the sensitivity of ET0 to the same climatic variables revealed significant differences among climates (Tabari and Talaee 2014). For example, Liang et al. (2008) examined the sensitivity of ET0 to air temperature, wind speed, relative humidity and sunshine hours in the Tao’er River basin of northeastern China. They showed that air temperature was the parameter of ET0 that was least sensitive, followed by wind speed, sunshine hours, and relative humidity. Zhang et al. (2010) conducted a sensitivity analysis of ET0 to temperature, vapor pressure deficit, wind speed, and available energy in Shiyang River basin of northwest China, and the results showed that air temperature was the least sensitive one as well, followed by wind speed, available energy, and vapor pressure deficit. The SC analysis in each soybean growth stage shows that SC values changed by the growth stage. This result was similar to the report by Liang et al. (2008) and Estévez et al. (2009). The variation patterns of SC of temperature and relative humidity were consistent with that of air temperature and relative humidity (Liang et al. 2008; Estévez et al. 2009). Solar radiation and wind speed did not show any clear relationship between their magnitude and the coefficient values (Liang et al. 2008; Estévez et al. 2009). The sensitivity of CWDI to the wind speed and sunshine hours shows opposite patterns in the present study, perhaps because of a decrease in the energetic term associated with an increase of the aerodynamic term, which led to the decrease of the SC for the shortwave radiation corresponding to an increase in the sensitivity coefficient for the wind speed (Gong et al. 2006). The sensitivity distribution reveals that SC changed spatially as well. The spatial variation of sensitivity of ET0 for the meteorological variables was related to the altitude, distance to the sea, and aridity conditions (Estévez et al. 2009). The lowest SC for temperature were found for stations located at high-altitude, high–wind speed stations with high wind sensitivity, and SC for relative humidity was related to the distance to the sea and the aridity conditions (Estévez et al. 2009). Liang et al. (2008) revealed that the sensitivity of ET0 for air temperature, relative humidity, and sunshine hours was distributed longitudinally, which was similar to the distribution pattern of three climate variables.

In general, the spatiotemporal assessment of soybean-related drought and sensitivity analysis in each growth stage not only provides important background and physical evidence for the influence of climate on soybean-related drought but also produces valuable information for agriculture planning and management in northeast China. The soybean seedling–branching stage and filling stages showed high CWDI, and the west area of northeast China showed high drought frequency. Precipitation was the most sensitive variable that influenced the CWDI. Thus, some drought-resistant management, especially related to water, can be practiced to mitigate soybean-related drought during the seedling–branching stage and filling stage in the west area of northeast China. People can build devices to collect water during the wet season and drain water during the dry season and can construct irrigation systems and divert water from areas with abundant water. Water-saving irrigation technologies, such as spray irrigation and micro-irrigation also can be used. In addition, mulching and conservation tillage are very effective drought-resistant technologies, drought-resistant soybean varieties and high-drought-resisting and water-preserving agents can also be adopted.

5. Conclusions

In this study, a comprehensive analysis of spatiotemporal variation of soybean-related drought and sensitivity to the meteorological variables in northeast China is presented. The following conclusions were drawn from this study: 1) During the soybean growing season in northeast China, precipitation and soybean water demand increase until the beginning of podding and then decrease gradually. The CWDI was higher during the seedling–branching, grain filling, and maturing stages, and the supply of precipitation could not meet the soybean water demands during these three stages. 2) Excessive drought occurred more often than light drought, medium drought, and severe drought in Heilongjiang, Jilin, and Liaoning. 3) The CWDI increased significantly during the filling stage in all three provinces from 1981 to 2010. 4) The spatial distribution of drought frequency exhibited a beltlike pattern; the most drought-prone area was found in the western part of northeast China, and the least drought-prone area was found in the southeastern part of northeast China. 5) Precipitation and relative humidity negatively affected the CWDI, while air temperature, wind speed, and number of sunshine hours positively affected the CWDI. 6) Precipitation was the most sensitive meteorological variable that influenced the CWDI, and wind speed was the least sensitive.

The soybean CWDI was determined using ET0, precipitation, and the crop coefficient. The factors affecting these variables were assumed to influence the CWDI as well. Evidence has shown that ET0 is affected by radiometric factors (Roderick and Farquhar 2002), aerodynamic factors (Xu et al. 2006; Rayner 2007; McVicar et al. 2008; Roderick et al. 2009a,b), and human activity (Tong et al. 2007), and the crop coefficient is affected by crop type, climate, soil evaporation and crop growth stages (Allen et al. 1998). In this study, only the climate was investigated. Further work is needed regarding other aspects to explain the spatial and temporal variations of soybean-related drought in northeast China.

Acknowledgments

This research was funded by The National Natural Science Foundation of China (41301594), the Special Fund for Agro-scientific Research in the Public Interest (201203031), and the National Science and Technology Basic Project (2007FY120100).

REFERENCES

  • Allen, R. G., L. S. Pereira, D. Raes, and M. Smith, 1998: Crop evapotranspiration—Guidelines for computing crop water requirement. FAO Irrigation and Drainage Paper 56, 15 pp. [Available online at http://www.fao.org/docrep/X0490E/X0490E00.htm.]

  • Cambardella, C. A., T. B. Moorman, J. M. Novak, T. B. Parkin, D. L. Karlen, R. F. Turco, and A. E. Konopka, 1994: Field-scale variability of soil properties in central Iowa soils. Soil. Sci. Soc. Amer. J., 58, 15011511, doi:10.2136/sssaj1994.03615995005800050033x.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Chen, C., C. Lei, A. Deng, C. Qian, W. Hoogmoed, and W. Zhang, 2011: Will higher minimum temperatures increase corn production in northeast China? An analysis of historical data over 1965–2008. Agric. For. Meteor., 151, 15801588, doi:10.1016/j.agrformet.2011.06.013.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Chen, S., Y. Shi, Y. Guo, and Y. X. Zheng, 2010: Temporal and spatial variation of annual mean air temperature in arid and semiarid region in northwest China over a recent 46 year period. J. Arid Land, 2, 8797, doi:10.3724/SP.J.1227.2010.00087.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Ding, Y., and Coauthors, 2007: History and Future Trend of Climate Change: China’s National Assessment Report on Climate Change. Science and Publishing House and Technology Press, 422 pp.

  • Espadafor, M., I. J. Lorite, P. Gavilán, and J. Berengena, 2011: An analysis of the tendency of reference evapotranspiration estimates and other climate variables during the last 45 years in southern Spain. Agric. Water Manage., 98, 10451061, doi:10.1016/j.agwat.2011.01.015.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Estévez, J., P. Gavilán, and J. Berengena, 2009: Sensitivity analysis of a Penman–Monteith type equation to estimate reference evapotranspiration in southern Spain. Hydrol. Processes, 23, 33423353, doi:10.1002/hyp.7439.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Gao, X., C. Wang, J. Zhang, and X. Xue, 2012: Crop water requirement and temporal-spatial variation of drought and flood disaster during growth stages for maize in northeast during past 50 years. Trans. Chin. Soc. Agric. Eng., 28, 101109.

    • Search Google Scholar
    • Export Citation
  • Gong, L., C. Xu, D. Chen, S. Halldin, and Y. D. Chen, 2006: Sensitivity of the Penman–Monteith reference evapotranspiration to key climatic variables in the Changjiang (Yangtze River) basin. J. Hydrol., 329, 620629, doi:10.1016/j.jhydrol.2006.03.027.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Gong, Y., W. S. Lu, and L. Tao, 2003: Spatial and temporal distribution of anomalous temperature/precipitation in spring and summer as well as their relationships to drought/flood in northeast of China. J. Nanjing Inst. Meteor., 26, 349357.

    • Search Google Scholar
    • Export Citation
  • Goyal, R. K., 2004: Sensitivity of evapotranspiration to global warming: A case study of arid zone of Rajasthan (India). Agric. Water Manage., 69, 111, doi:10.1016/j.agwat.2004.03.014.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Guenang, G. M., and F. M. Kamga, 2014: Computation of the standardized precipitation index (SPI) and its use to assess drought occurrences in Cameroon over recent decades. J. Appl. Meteor. Climatol., 53, 23102324, doi:10.1175/JAMC-D-14-0032.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Huang, W. H., and Coauthors, 2009: Analysis of spatio-temporal characteristic on seasonal drought of spring maize based on crop water deficit index. Trans. Chin. Soc. Agric. Eng., 25, 2834.

    • Search Google Scholar
    • Export Citation
  • IPCC, 2007: Summary for policymakers. Climate Change 2007: The Physical Science Basis, S. Solomon et al., Eds., Cambridge University Press, 1–18.

    • Crossref
    • Export Citation
  • IPCC, 2013: Summary for policymakers. Climate Change 2013: The Physical Science Basis, T. F. Stocker et al., Eds., Cambridge University Press, 1–29.

  • Irmak, S., J. O. Payero, D. L. Martin, A. Irmak, and T. A. Howell, 2006: Sensitivity analyses and sensitivity coefficients of standardized daily ASCE-Penman–Monteith equation. J. Irrig. Drain. Eng., 132, 564578, doi:10.1061/(ASCE)0733-9437(2006)132:6(564).

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Kendall, M. G., 1975: Rank Correlation Measures. Charles Griffin, 202 pp.

  • Keyantash, J., and A. J. Dracup, 2002: The quantification of drought: An evaluation of drought indices. Bull. Amer. Meteor. Soc., 83, 11671180, doi:10.1175/1520-0477(2002)083<1191:TQODAE>2.3.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Kim, D., H. Byun, K. Choi, and S. Oh, 2011: A spatiotemporal analysis of historical droughts in Korea. J. Appl. Meteor. Climatol., 50, 18951912, doi:10.1175/2011JAMC2664.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Lee, S., H. Byun, and H. L. Tanaka, 2012: Spatiotemporal characteristics of drought occurrences over Japan. J. Appl. Meteor. Climatol., 51, 10871098, doi:10.1175/JAMC-D-11-0157.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Li, R., A. Tsunekawa, and M. Tsubo, 2014: Index-based assessment of agricultural drought in a semi-arid region of Inner Mongolia, China. J. Arid Land, 6, 315, doi:10.1007/s40333-013-0193-8.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Liang, L., L. Li, L. Zhang, J. Li, and B. Li, 2008: Sensitivity of Penman-Monteith reference crop evapotranspiration in Tao’er River basin of northeastern China. Chin. Geogr. Sci., 18, 340347, doi:10.1007/s11769-008-0340-x.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Liu, Z. J., X. G. Yang, W. F. Wang, K. N. Li, and X. Y. Zhang, 2009: Characteristics of agricultural climate resources in three provinces of northeast China under global climate change. Chin. J. Appl. Ecol., 20, 21992206.

    • Search Google Scholar
    • Export Citation
  • Lv, H., Y. Zhang, J. Wang, S. Zhang, X. Lou, and Y. Zhang, 2009: Classification of agricultural drought category. General Administration of Quality Supervision, Inspection and Quarantine of the People’s Republic of China, Beijing, China, 44 pp.

  • Mann, H. B., 1945: Nonparametric tests against trend. Econometrica, 13, 245259, doi:10.2307/1907187.

  • McCabe, G. J., and D. M. Wolock, 2015: Variability and trends in global drought. Earth Space Sci., 2, 223228, doi:10.1002/2015EA000100.

  • McKenney, M. S., and N. J. Rosenberg, 1993: Sensitivity of some potential evapotranspiration estimation methods to climate change. Agric. For. Meteor., 64, 81110, doi:10.1016/0168-1923(93)90095-Y.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • McVicar, T. R., T. G. Van Niel, L. T. Li, M. L. Roderick, D. P. Rayner, L. Ricciardulli, and R. J. Donohue, 2008: Wind speed climatology and trends for Australia, 1975–2006: Capturing the stilling phenomenon and comparison with near-surface analysis output. Geophys. Res. Lett., 35, 16, doi:10.1029/2008GL035627.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Min, S. K., X. Zhang, F. W. Zwiers, and G. C. Hegerl, 2011: Human contribution to more-intense precipitation extremes. Nature, 470, 378381, doi:10.1038/nature09763.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Mitchell, J. M., B. Dzerdzeevskii, H. Flohn, W. L. Hofmeyr, H. H. Lamb, K. N. Rao, and C. C. Wallen, 1966: Climate change. WMO Tech. Note 79, 79 pp.

  • Monteith, J. L., 1965: Evaporation and the environment. The State and Movement of Water in Living Organism, G. E. Fogg, Ed., Cambridge University Press, 432 pp.

  • Ning, L., and R. S. Bradley, 2015: Snow occurrence changes over the central and eastern United States under future warming scenarios. Sci. Rep., 5, 17073, doi:10.1038/srep17073.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Ning, L., and R. S. Bradley, 2016: NAO and PNA influences on winter temperature and precipitation over the eastern United States in CMIP5 GCMs. Climate Dyn., 46, 12571276, doi:10.1007/s00382-015-2643-9.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Rana, G., and N. Katerji, 1998: A measurement based sensitivity analysis of the Penman–Monteith actual evapotranspiration model for crops of different height and in contrasting water status. Theor. Appl. Climatol., 60, 141149, doi:10.1007/s007040050039.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Rayner, D. P., 2007: Wind run changes: The dominant factor affecting pan evaporation trends in Australia. J. Climate, 20, 33793394, doi:10.1175/JCLI4181.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Roderick, M. L., and G. D. Farquhar, 2002: The cause of decreased pan evaporation over the past 50 years. Science, 298, 14101411, doi:10.1126/science.1075390-a.

    • Search Google Scholar
    • Export Citation
  • Roderick, M. L., M. T. Hobbins, and G. D. Farquhar, 2009a: Pan evaporation trends and the terrestrial water balance. I. Principles and observations. Geogr. Compass, 3, 746760, doi:10.1111/j.1749-8198.2008.00213.x.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Roderick, M. L., M. T. Hobbins, and G. D. Farquhar, 2009b: Pan evaporation trends and the terrestrial water balance. II. Energy balance and interpretation. Geogr. Compass, 3, 761780, doi:10.1111/j.1749-8198.2008.00214.x.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Sheffield, J., and E. F. Wood, 2008: Global trends and variability in soil moisture and drought characteristics, 1950–2000, from observation-driven simulations of the terrestrial hydrologic cycle. J. Climate, 21, 432458, doi:10.1175/2007JCLI1822.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Sonali, P., and D. N. Kumar, 2013: Review of trend detection methods and their application to detect temperature changes in India. J. Hydrol., 476, 212227, doi:10.1016/j.jhydrol.2012.10.034.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Sui, Y., W. H. Huang, X. G. Yang, and M. S. Li, 2012: Characteristics and adaption of seasonal drought in southern China under the background of global climate change. II. Spatiotemporal characteristics of drought for wintering grain- and oil crops based on crop water deficit index. Yingyong Shengtai Xuebao, 23, 24672476.

    • Search Google Scholar
    • Export Citation
  • Sun, F., S. Yang, and P. Chen, 2005: Climatic warming-drying trend in northeastern China during the last 44 years and its effects. Shengtaixue Zazhi, 24, 751755.

    • Search Google Scholar
    • Export Citation
  • Sun, F., J. Yuan, and S. Lu, 2006: The change and test of climate in northeast China over last 100 years. Climate Environ. Res., 11, 101107.

    • Search Google Scholar
    • Export Citation
  • Sun, F., S. Yang, and G. Ren, 2007: Decade variations of precipitation event frequency, intensity and duration in the northeast China. J. Appl. Meteor. Sci., 18, 610618.

    • Search Google Scholar
    • Export Citation
  • Tabari, H., and P. H. Talaee, 2014: Sensitivity of evapotranspiration to climatic change in different climates. Global Planet. Change, 115, 1623, doi:10.1016/j.gloplacha.2014.01.006.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Tong, L., S. Kang, L. Zhang, and L. Zhang, 2007: Temporal and spatial variations of evapotranspiration for spring wheat in the Shiyang River basin in northwest China. Agric. Water Manage., 87, 241250, doi:10.1016/j.agwat.2006.07.013.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wang, H., and Coauthors, 2012: Extreme climate in China: Facts, simulation and projection. Meteor. Z., 21, 279304, doi:10.1127/0941-2948/2012/0330.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wen, K., and Coauthors, 2005: China’s Meteorological Disaster Encyclopedia. Meteorological Press, 322 pp.

  • Xie, W. J., X. G. Yang, J. Yang, L. M. Liu, Q. Ye, C. Y. Dong, Z. J. Liu, and J. Zhao, 2014: Spatio-temporal characteristics of drought for soybean under climate change in the three provinces of northeast China. Acta Ecol. Sin., 34, 62326243.

    • Search Google Scholar
    • Export Citation
  • Xu, C., L. Gong, T. Jiang, D. Chen, and V. P. Singh, 2006: Analysis of spatial distribution and temporal trend of reference evapotranspiration and pan evaporation in Changjiang (Yangtze River) catchment. J. Hydrol., 327, 8193, doi:10.1016/j.jhydrol.2005.11.029.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Xue, Q., 2013: Analysis on the change of 30 year’s soybean areas, production and yield in China and northeast China. Zhongguo Nongxue Tongbao, 29, 102106.

    • Search Google Scholar
    • Export Citation
  • Yan, C. J., W. B. Wang, X. J. Tu, C. L. Wang, L. J. Zhang, Q. Du, and S. H. Song, 2013: Effect of drought stress at different growth stage on yield and root characteristics of soybean. Dadou Kexue, 32, 5962.

    • Search Google Scholar
    • Export Citation
  • Yang, X., L. Ren, Y. Liu, D. Jiao, and S. Jiang, 2014: Hydrological response to land use and land cover changes in a sub-watershed of west Liaohe River basin, China. J. Arid Land, 6, 678689, doi:10.1007/s40333-014-0026-4.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Yin, X. G., J. E. Olesen, M. Wang, I. Öztürk, and F. Chen, 2016: Climate effects on crop yields in the northeast farming region of China during 1961–2010. J. Agric. Sci., 154, 11901208, doi:10.1017/S0021859616000149.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Zhang, M., J. He, B. Wang, S. Wang, S. Li, W. Liu, and X. Ma, 2013: Extreme drought changes in southwest China from 1960 to 2009. J. Geogr. Sci., 23, 316, doi:10.1007/s11442-013-0989-7.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Zhang, S., Y. Zhang, R. Ji, F. Cai, and J. Wu, 2011: Analysis of spatio-temporal characteristics of drought for maize in northeast China. Agric. Res. Arid Areas, 29, 231236.

    • Search Google Scholar
    • Export Citation
  • Zhang, X., S. Kang, L. Zhang, and J. Liu, 2010: Spatial variation of climatology monthly crop reference evapotranspiration and sensitivity coefficients in Shiyang River basin of northwest China. Agric. Water Manage., 97, 15061516, doi:10.1016/j.agwat.2010.05.004.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Zhang, Y., H. Lu, and S. Li, 2008: Applicability of crop water deficit index in agricultural drought monitoring. Mater. Sci. Technol., 36, 596600.

    • Search Google Scholar
    • Export Citation
  • Zhao, C., G. Ren, and Y. Zhang, 2009: Climate change of the northeast China over the past 50 years. Ganhanqu Ziyuan Yu Huanjing, 23, 2530.

    • Search Google Scholar
    • Export Citation
Save