Acquiring spatiotemporal patterns of phenological information and its drivers is essential for understanding the response of crops to climate change and implementing adaptation measures.However,current approaches to o...Acquiring spatiotemporal patterns of phenological information and its drivers is essential for understanding the response of crops to climate change and implementing adaptation measures.However,current approaches to obtain phenology and analyse its drivers have deficiencies such as sparse observations,excessive dependence of remote sensing inversion on sensors,and inevitable difficulties in upscaling site-based crop models into larger regions.Based on the Wang-Engel temperature response function,we improved the Crop Estimation through Resource and Environment Synthesis-Wheat(CERES-Wheat)model.First,we calibrated the model at the regional scale and evaluated its performance.Furthermore,the spatiotemporal changes in winter wheat phenology in China from 2000 to 2015 were analysed.The results showed that the improved model significantly enhanced the simulation accuracy of the anthesis and maturity dates by averages of 13%and 12%in most planting areas,especially in the Yunnan-Guizhou Plateau(YG)with improvements of 26%and 28%.The simulated phenology of winter wheat grown in a colder environment(e.g.,the average temperatures during the vegetative growth period range from 0 to 5℃ and from 15 to 20°C,and the reproductive growth period ranges from 10 to 15°C)also notably improved.These results confirmed that the original temperature response function indeed had limitations.Further analyses revealed that the key phenological dates and growth periods over the past 16 years were dominantly advanced and shortened.Specifically,the anthesis date,vegetative growth period(VGP),and reproductive growth period(RGP)indicated obviously spatial characteristics.For example,the anthesis date and VGP in the North China Plain(NCP)and the Middle-Lower Yangtze Plain(YZ)and the RGP in northwestern China(NW)showed opposite trends of delay and prolongation as comparing with the dominant patterns.Sensitivity analysis indicated that the key phenological dates and growth periods were advanced and shortened as the minimum(T_(min))and maximum temperatures(T_(max))rose,while they were postponed and prolonged with the increased precipitation.However,their responses to solar radiation did not show spatial consistency.Additionally,we found that the sensitivity of phenology to climatic factors differed across subregions.In particular,phenology in southwestern China and YG was more sensitive to T_(min),T_(max),and solar radiation than in the NCP and NW.Moreover,the sensitivity to precipitation in NW was higher than that in YZ.Totally,the improved crop model could provide more refined spatial characteristics of phenology at a large scale and benefit to explore its drivers more objectively.Furthermore,our results highlight that different planting areas should adopt suitable adaptation measures to cope with climate change impacts.Ultimately,the improved model is promising to enhance the accuracy of yield prediction and provide powerful tools for assessing regional climate change impact and adaptability.展开更多
基金This work was supported by the National Natural Science Foundation of China(Grant Nos.41977405,42061144003).
文摘Acquiring spatiotemporal patterns of phenological information and its drivers is essential for understanding the response of crops to climate change and implementing adaptation measures.However,current approaches to obtain phenology and analyse its drivers have deficiencies such as sparse observations,excessive dependence of remote sensing inversion on sensors,and inevitable difficulties in upscaling site-based crop models into larger regions.Based on the Wang-Engel temperature response function,we improved the Crop Estimation through Resource and Environment Synthesis-Wheat(CERES-Wheat)model.First,we calibrated the model at the regional scale and evaluated its performance.Furthermore,the spatiotemporal changes in winter wheat phenology in China from 2000 to 2015 were analysed.The results showed that the improved model significantly enhanced the simulation accuracy of the anthesis and maturity dates by averages of 13%and 12%in most planting areas,especially in the Yunnan-Guizhou Plateau(YG)with improvements of 26%and 28%.The simulated phenology of winter wheat grown in a colder environment(e.g.,the average temperatures during the vegetative growth period range from 0 to 5℃ and from 15 to 20°C,and the reproductive growth period ranges from 10 to 15°C)also notably improved.These results confirmed that the original temperature response function indeed had limitations.Further analyses revealed that the key phenological dates and growth periods over the past 16 years were dominantly advanced and shortened.Specifically,the anthesis date,vegetative growth period(VGP),and reproductive growth period(RGP)indicated obviously spatial characteristics.For example,the anthesis date and VGP in the North China Plain(NCP)and the Middle-Lower Yangtze Plain(YZ)and the RGP in northwestern China(NW)showed opposite trends of delay and prolongation as comparing with the dominant patterns.Sensitivity analysis indicated that the key phenological dates and growth periods were advanced and shortened as the minimum(T_(min))and maximum temperatures(T_(max))rose,while they were postponed and prolonged with the increased precipitation.However,their responses to solar radiation did not show spatial consistency.Additionally,we found that the sensitivity of phenology to climatic factors differed across subregions.In particular,phenology in southwestern China and YG was more sensitive to T_(min),T_(max),and solar radiation than in the NCP and NW.Moreover,the sensitivity to precipitation in NW was higher than that in YZ.Totally,the improved crop model could provide more refined spatial characteristics of phenology at a large scale and benefit to explore its drivers more objectively.Furthermore,our results highlight that different planting areas should adopt suitable adaptation measures to cope with climate change impacts.Ultimately,the improved model is promising to enhance the accuracy of yield prediction and provide powerful tools for assessing regional climate change impact and adaptability.
