This study assesses the suitability of convolutional neural networks(CNNs) for downscaling precipitation over East Africa in the context of seasonal forecasting. To achieve this, we design a set of experiments that co...This study assesses the suitability of convolutional neural networks(CNNs) for downscaling precipitation over East Africa in the context of seasonal forecasting. To achieve this, we design a set of experiments that compare different CNN configurations and deployed the best-performing architecture to downscale one-month lead seasonal forecasts of June–July–August–September(JJAS) precipitation from the Nanjing University of Information Science and Technology Climate Forecast System version 1.0(NUIST-CFS1.0) for 1982–2020. We also perform hyper-parameter optimization and introduce predictors over a larger area to include information about the main large-scale circulations that drive precipitation over the East Africa region, which improves the downscaling results. Finally, we validate the raw model and downscaled forecasts in terms of both deterministic and probabilistic verification metrics, as well as their ability to reproduce the observed precipitation extreme and spell indicator indices. The results show that the CNN-based downscaling consistently improves the raw model forecasts, with lower bias and more accurate representations of the observed mean and extreme precipitation spatial patterns. Besides, CNN-based downscaling yields a much more accurate forecast of extreme and spell indicators and reduces the significant relative biases exhibited by the raw model predictions. Moreover, our results show that CNN-based downscaling yields better skill scores than the raw model forecasts over most portions of East Africa. The results demonstrate the potential usefulness of CNN in downscaling seasonal precipitation predictions over East Africa,particularly in providing improved forecast products which are essential for end users.展开更多
Based on the lightning observation data from the Fengyun-4A(FY-4A)Lightning Mapping Imager(FY-4A/LMI)and the Lightning Imaging Sensor(LIS)on the International Space Station(ISS),we extract the“event”type data as the...Based on the lightning observation data from the Fengyun-4A(FY-4A)Lightning Mapping Imager(FY-4A/LMI)and the Lightning Imaging Sensor(LIS)on the International Space Station(ISS),we extract the“event”type data as the lightning detection results.These observations are then compared with the cloud-to-ground(CG)lightning observation data from the China Meteorological Administration.This study focuses on the characteristics of lightning activity in Southeast China,primarily in Jiangxi Province and its adjacent areas,from April to September,2017–2022.In addition,with the fifth-generation European Centre for Medium-Range Weather Forecasts reanalysis data,we further delved into the potential factors influencing the distribution and variations in lightning activity and their primary related factors.Our findings indicate that the lightning frequency and density of the FY-4A/LMI,ISS-LIS and CG data are higher in southern and central Jiangxi,central Fujian Province,and western and central Guangdong Province,while they tend to be lower in eastern Hunan Province.In general,the high-value areas of lightning density for the FY-4A/LMI are located in inland mountainous areas.The lower the latitude is,the higher the CG lightning density is.High-value areas of the CG lightning density are more likely to be located in eastern Fujian and southeastern Zhejiang Province.However,the high-value areas of lightning density for the ISS-LIS are more dispersed,with a scattered distribution in inland mountainous areas and along the coast of eastern Fujian.Thus,the mountainous terrain is closely related to the high-value areas of the lightning density.The locations of the high-value areas of the lightning density for the FY-4A/LMI correspond well with those for the CG observations,and the seasonal variations are also consistent.In contrast,the distribution of the high-value areas of the lightning density for the ISS-LIS is more dispersed.The positions of the peak frequency of the FY-4A/LMI lightning and CG lightning contrast with local altitudes,primarily located at lower altitudes or near mountainsides.K-index and convective available potential energy(CAPE)can better reflect the local boundary layer conditions,where the lightning density is higher and lightning seasonal variations are apparent.There are strong correlations in the annual variations between the dew-point temperature(Td)and CG lightning frequency,and the monthly variations of the dew-point temperature and CAPE are also strongly correlated with monthly variations of CG lightning,while they are weakly correlated with the lightning frequency for the FY-4A/LMI and ISS-LIS.This result reflects that the CAPE shows a remarkable effect on the CG lightning frequency during seasonal transitions.展开更多
Few studies have investigated the spatial patterns of the air temperature urban heat island(AUHI)and its controlling factors.In this study,the data generated by an urban climate model were used to investigate the spat...Few studies have investigated the spatial patterns of the air temperature urban heat island(AUHI)and its controlling factors.In this study,the data generated by an urban climate model were used to investigate the spatial variations of the AUHI across China and the underlying climate and ecological drivers.A total of 355 urban clusters were used.We performed an attribution analysis of the AUHI to elucidate the mechanisms underlying its formation.The results show that the midday AUHI is negatively correlated with climate wetness(humid:0.34 K;semi-humid:0.50 K;semi-arid:0.73 K).The annual mean midnight AUHI does not show discernible spatial patterns,but is generally stronger than the midday AUHI.The urban–rural difference in convection efficiency is the largest contributor to the midday AUHI in the humid(0.32±0.09 K)and the semi-arid(0.36±0.11 K)climate zones.The release of anthropogenic heat from urban land is the dominant contributor to the midnight AUHI in all three climate zones.The rural vegetation density is the most important driver of the daytime and nighttime AUHI spatial variations.A spatial covariance analysis revealed that this vegetation influence is manifested mainly through its regulation of heat storage in rural land.展开更多
The representation of the Arctic stratospheric circulation and the quasi-biennial oscillation(QBO)during the period 1981–2019 in a 40-yr Chinese global reanalysis dataset(CRA-40)is evaluated by comparing two widely u...The representation of the Arctic stratospheric circulation and the quasi-biennial oscillation(QBO)during the period 1981–2019 in a 40-yr Chinese global reanalysis dataset(CRA-40)is evaluated by comparing two widely used reanalysis datasets,ERA-5 and MERRA-2.CRA-40 demonstrates a comparable performance with ERA-5 and MERRA-2 in characterizing the winter and spring circulation in the lower and middle Arctic stratosphere.Specifically,differences in the climatological polar-mean temperature and polar night jet among the three reanalyses are within±0.5 K and±0.5 m s^(–1),respectively.The onset dates of the stratospheric sudden warming and stratospheric final warming events at 10 hPa in CRA-40,together with the dynamics and circulation anomalies during the onset process of warming events,are nearly identical to the other two reanalyses with slight differences.By contrast,the CRA-40 dataset demonstrates a deteriorated performance in describing the QBO below 10 hPa compared to the other two reanalysis products,manifested by the larger easterly biases of the QBO index,the remarkably weaker amplitude of the QBO,and the weaker wavelet power of the QBO period.Such pronounced biases are mainly concentrated in the period 1981–98 and largely reduced by at least 39%in 1999–2019.Thus,particular caution is needed in studying the QBO based on CRA-40.All three reanalyses exhibit greater disagreement in the upper stratosphere compared to the lower and middle stratosphere for both the polar region and the tropics.展开更多
The present study explored how the Indian Ocean Dipole (IOD) influences October-November-December (OND) rainfall over Tanzania in recent decade following the 2011 abrupt change. The study spans 50 years, from 1973 to ...The present study explored how the Indian Ocean Dipole (IOD) influences October-November-December (OND) rainfall over Tanzania in recent decade following the 2011 abrupt change. The study spans 50 years, from 1973 to 2022. Notable abrupt changes were observed in 1976 and 2011, leading us to divide our study into two periods: 1976-2010 and 2011-2022, allowing for a close investigation into the existing relationship between OND IOD and OND rainfall and their associated large-scale atmospheric circulations. It was found that the relationship between OND IOD and OND rainfall strengthened, with the correlation changed from +0.73 during 1976-2010 to +0.81 during 2011-2022. Further investigation revealed that, during 1976-2010, areas that received above- normal rainfall during positive IOD experienced below-normal during 2011- 2022 and vice versa. The same pattern relationship was observed for negative IOD. Spatial analysis demonstrates that the percentage departure of rainfall across the region mirrors the standardized rainfall anomalies. The study highlights that the changing relationship between OND IOD and OND rainfall corresponds to the east-west shift of Walker circulation, as well as the north-south shift of Hadley circulation. Analysis of sea surface temperature (SST) indicates that both positive and negative IOD events strengthened during 2011-2022 compared to 1976-2010. Close monitoring of this relationship across different timescales could be useful for updating OND rainfall seasonal forecasts in Tanzania, serving as a tool for reducing socio-economic impacts.展开更多
Understanding the characteristics of extreme rainfall is crucial for effective flood management planning, as it enables the incorporation of insights from past extreme rainfall patterns and their spatiotemporal distri...Understanding the characteristics of extreme rainfall is crucial for effective flood management planning, as it enables the incorporation of insights from past extreme rainfall patterns and their spatiotemporal distribution. This work investigated the changes in the frequency and pattern of extreme rainfall over Uganda, using daily datasets sourced from Climate Hazard Group InfraRed Precipitation with Station (CHIRPS-v2) for the period 1981 to 2022. The study utilized the extreme weather Indices provided by the Expert Team on Climate Change Detection and Indices (ETCCDI). Attention was directed towards September to November (SON) rainfall season with precise analysis of four indices (Rx1day, Rx5day, R95p, and R99p). The Sequential Mann-Kendall (SQMK) non-parametric test was applied to identify abrupt changes in SON extreme rainfall trends. Results showed that October consistently recorded the highest count of extreme rainfall days across all four indices. The long-term analysis revealed fluctuations in extreme rainfall events across years, with certain periods exhibiting heightened intensity. The analysis portrayed a shift in the decadal variations and region-specific distribution of extreme rainfall, with Eastern Uganda and areas around Lake Victoria standing out compared to other regions. The findings further revealed an increase in extreme rainfall for all indices in the recent decade (2011-2022) with 2019/2020 standing out as the extreme years of SON for the study period. While trendlines suggested a slight increase in intense daily rainfall events, the SQMK tests revealed statistical significance in the trend of prolonged periods of intense daily rainfall. This study contributes to the understanding of the spatiotemporal variability and trends of extreme rainfall events over Uganda during the SON season, which is crucial for the assessment of climate change impacts and adaptation strategies. It provides valuable information for seasonal extreme rainfall forecasting, development of early warning systems, flood risk management, and disaster preparedness plans.展开更多
This paper presents an attempt at assimilating clear-sky FY-4A Advanced Geosynchronous Radiation Imager(AGRI)radiances from two water vapor channels for the prediction of three landfalling typhoon events over the West...This paper presents an attempt at assimilating clear-sky FY-4A Advanced Geosynchronous Radiation Imager(AGRI)radiances from two water vapor channels for the prediction of three landfalling typhoon events over the West Pacific Ocean using the 3DVar data assimilation(DA)method along with the WRF model.A channel-sensitive cloud detection scheme based on the particle filter(PF)algorithm is developed and examined against a cloud detection scheme using the multivariate and minimum residual(MMR)algorithm and another traditional cloud mask–dependent cloud detection scheme.Results show that both channel-sensitive cloud detection schemes are effective,while the PF scheme is able to reserve more pixels than the MMR scheme for the same channel.In general,the added value of AGRI radiances is confirmed when comparing with the control experiment without AGRI radiances.Moreover,it is found that the analysis fields of the PF experiment are mostly improved in terms of better depicting the typhoon,including the temperature,moisture,and dynamical conditions.The typhoon track forecast skill is improved with AGRI radiance DA,which could be explained by better simulating the upper trough.The impact of assimilating AGRI radiances on typhoon intensity forecasts is small.On the other hand,improved rainfall forecasts from AGRI DA experiments are found along with reduced errors for both the thermodynamic and moisture fields,albeit the improvements are limited.展开更多
为了准确监测和客观评估稻纵卷叶螟对水稻生长发育和产量形成的危害,利用ASD Field Spec3地物波谱仪和SPAD-502叶绿素仪分别采集控制大田试验(2015年和2019年)和自然大田试验(2020年)在各生育期(拔节期、孕穗期、灌浆期、成熟期)水稻的...为了准确监测和客观评估稻纵卷叶螟对水稻生长发育和产量形成的危害,利用ASD Field Spec3地物波谱仪和SPAD-502叶绿素仪分别采集控制大田试验(2015年和2019年)和自然大田试验(2020年)在各生育期(拔节期、孕穗期、灌浆期、成熟期)水稻的冠层高光谱数据和SPAD值,调查采集样点的虫量和水稻卷叶率,对比分析两种试验中稻纵卷叶螟的虫害发生特征、水稻冠层光谱特征和水稻生理生态参数特征,建立基于高光谱参数的水稻受稻纵卷叶螟危害的生理生态参数估算模型。结果表明,(1)两种试验的水稻SPAD值和冠层的红边至近红外波段的反射率均随着稻纵卷叶螟虫害程度的加重而降低,而可见光波段的反射率则相反;(2)自然大田试验的SPAD值和红光至近红外波段的冠层反射率在水稻生长发育前期要显著低于控制大田试验,而到了后期则反而要略高于控制大田试验;(3)综合分析筛选出自然大田试验和控制大田试验中的多个虫害特征参数和植被指数分别构建出了SPAD的单因子和多因子估算模型,各模型均达到了较好的估算效果,在单因子模型中EVI的二项式函数模拟效果最好,而多因子线性回归估测模型的模拟效果优于所有的单因子模型;(4)通过2021年对这些模型的应用检验发现:这些模型中基于虫量、卷叶率、OSAVI、EVI和DVI的单因子估算模型的SPAD估算值与实测值拟合度很高,其Rv 2均超过了0.8,达到了比较理想的估算效果,这为稻纵卷叶螟危害下的水稻SPAD值估测提供了一种精度较高且可行的估算方法。展开更多
基金supported by the National Key Research and Development Program of China (Grant No.2020YFA0608000)the National Natural Science Foundation of China (Grant No. 42030605)the High-Performance Computing of Nanjing University of Information Science&Technology for their support of this work。
文摘This study assesses the suitability of convolutional neural networks(CNNs) for downscaling precipitation over East Africa in the context of seasonal forecasting. To achieve this, we design a set of experiments that compare different CNN configurations and deployed the best-performing architecture to downscale one-month lead seasonal forecasts of June–July–August–September(JJAS) precipitation from the Nanjing University of Information Science and Technology Climate Forecast System version 1.0(NUIST-CFS1.0) for 1982–2020. We also perform hyper-parameter optimization and introduce predictors over a larger area to include information about the main large-scale circulations that drive precipitation over the East Africa region, which improves the downscaling results. Finally, we validate the raw model and downscaled forecasts in terms of both deterministic and probabilistic verification metrics, as well as their ability to reproduce the observed precipitation extreme and spell indicator indices. The results show that the CNN-based downscaling consistently improves the raw model forecasts, with lower bias and more accurate representations of the observed mean and extreme precipitation spatial patterns. Besides, CNN-based downscaling yields a much more accurate forecast of extreme and spell indicators and reduces the significant relative biases exhibited by the raw model predictions. Moreover, our results show that CNN-based downscaling yields better skill scores than the raw model forecasts over most portions of East Africa. The results demonstrate the potential usefulness of CNN in downscaling seasonal precipitation predictions over East Africa,particularly in providing improved forecast products which are essential for end users.
基金supported by the National Natural Science Foundation of China[grant number 42088101] and the National Natural Science Foundation of China[grant number 42005020].
基金National Natural Science Foundation of China(42175014,42205137)Open Research Fund of Institute of Meteorological Technology Innovation,Nanjing(BJG202202)+3 种基金Joint Research Project of Typhoon Research,Shanghai Typhoon Institute,China Meteorological Administration(TFJJ202209)Innovation Development Project of China Meteorological Administration(CXFZ2023P001)Open Project of KLME&CIC-FEMD(KLME202311)Jiangxi MDIA-ASI Fund。
文摘Based on the lightning observation data from the Fengyun-4A(FY-4A)Lightning Mapping Imager(FY-4A/LMI)and the Lightning Imaging Sensor(LIS)on the International Space Station(ISS),we extract the“event”type data as the lightning detection results.These observations are then compared with the cloud-to-ground(CG)lightning observation data from the China Meteorological Administration.This study focuses on the characteristics of lightning activity in Southeast China,primarily in Jiangxi Province and its adjacent areas,from April to September,2017–2022.In addition,with the fifth-generation European Centre for Medium-Range Weather Forecasts reanalysis data,we further delved into the potential factors influencing the distribution and variations in lightning activity and their primary related factors.Our findings indicate that the lightning frequency and density of the FY-4A/LMI,ISS-LIS and CG data are higher in southern and central Jiangxi,central Fujian Province,and western and central Guangdong Province,while they tend to be lower in eastern Hunan Province.In general,the high-value areas of lightning density for the FY-4A/LMI are located in inland mountainous areas.The lower the latitude is,the higher the CG lightning density is.High-value areas of the CG lightning density are more likely to be located in eastern Fujian and southeastern Zhejiang Province.However,the high-value areas of lightning density for the ISS-LIS are more dispersed,with a scattered distribution in inland mountainous areas and along the coast of eastern Fujian.Thus,the mountainous terrain is closely related to the high-value areas of the lightning density.The locations of the high-value areas of the lightning density for the FY-4A/LMI correspond well with those for the CG observations,and the seasonal variations are also consistent.In contrast,the distribution of the high-value areas of the lightning density for the ISS-LIS is more dispersed.The positions of the peak frequency of the FY-4A/LMI lightning and CG lightning contrast with local altitudes,primarily located at lower altitudes or near mountainsides.K-index and convective available potential energy(CAPE)can better reflect the local boundary layer conditions,where the lightning density is higher and lightning seasonal variations are apparent.There are strong correlations in the annual variations between the dew-point temperature(Td)and CG lightning frequency,and the monthly variations of the dew-point temperature and CAPE are also strongly correlated with monthly variations of CG lightning,while they are weakly correlated with the lightning frequency for the FY-4A/LMI and ISS-LIS.This result reflects that the CAPE shows a remarkable effect on the CG lightning frequency during seasonal transitions.
基金supported by the National Key R&D Program of China (Grant No.2019YFA0607202)the National Natural Science Foundation of China (Grant Nos. 42021004 and 42005143)+2 种基金support by the Postgraduate Research&Practice Innovation Program of Jiangsu Province (Grant No. KYCX21_0978)support by the Open Research Fund Program of the Key Laboratory of Urban Meteorology,China Meteorological Administration (Grant No. LUM-2023-12)the 333 Project of Jiangsu Province (Grant No. BRA2022023)
文摘Few studies have investigated the spatial patterns of the air temperature urban heat island(AUHI)and its controlling factors.In this study,the data generated by an urban climate model were used to investigate the spatial variations of the AUHI across China and the underlying climate and ecological drivers.A total of 355 urban clusters were used.We performed an attribution analysis of the AUHI to elucidate the mechanisms underlying its formation.The results show that the midday AUHI is negatively correlated with climate wetness(humid:0.34 K;semi-humid:0.50 K;semi-arid:0.73 K).The annual mean midnight AUHI does not show discernible spatial patterns,but is generally stronger than the midday AUHI.The urban–rural difference in convection efficiency is the largest contributor to the midday AUHI in the humid(0.32±0.09 K)and the semi-arid(0.36±0.11 K)climate zones.The release of anthropogenic heat from urban land is the dominant contributor to the midnight AUHI in all three climate zones.The rural vegetation density is the most important driver of the daytime and nighttime AUHI spatial variations.A spatial covariance analysis revealed that this vegetation influence is manifested mainly through its regulation of heat storage in rural land.
基金supported by the National Natural Science Foundation of China (Grant Nos. 41975048, 42030605, and 42175069)the Natural Science Foundation of Jiangsu Province (Grant No.BK20191404)
文摘The representation of the Arctic stratospheric circulation and the quasi-biennial oscillation(QBO)during the period 1981–2019 in a 40-yr Chinese global reanalysis dataset(CRA-40)is evaluated by comparing two widely used reanalysis datasets,ERA-5 and MERRA-2.CRA-40 demonstrates a comparable performance with ERA-5 and MERRA-2 in characterizing the winter and spring circulation in the lower and middle Arctic stratosphere.Specifically,differences in the climatological polar-mean temperature and polar night jet among the three reanalyses are within±0.5 K and±0.5 m s^(–1),respectively.The onset dates of the stratospheric sudden warming and stratospheric final warming events at 10 hPa in CRA-40,together with the dynamics and circulation anomalies during the onset process of warming events,are nearly identical to the other two reanalyses with slight differences.By contrast,the CRA-40 dataset demonstrates a deteriorated performance in describing the QBO below 10 hPa compared to the other two reanalysis products,manifested by the larger easterly biases of the QBO index,the remarkably weaker amplitude of the QBO,and the weaker wavelet power of the QBO period.Such pronounced biases are mainly concentrated in the period 1981–98 and largely reduced by at least 39%in 1999–2019.Thus,particular caution is needed in studying the QBO based on CRA-40.All three reanalyses exhibit greater disagreement in the upper stratosphere compared to the lower and middle stratosphere for both the polar region and the tropics.
文摘The present study explored how the Indian Ocean Dipole (IOD) influences October-November-December (OND) rainfall over Tanzania in recent decade following the 2011 abrupt change. The study spans 50 years, from 1973 to 2022. Notable abrupt changes were observed in 1976 and 2011, leading us to divide our study into two periods: 1976-2010 and 2011-2022, allowing for a close investigation into the existing relationship between OND IOD and OND rainfall and their associated large-scale atmospheric circulations. It was found that the relationship between OND IOD and OND rainfall strengthened, with the correlation changed from +0.73 during 1976-2010 to +0.81 during 2011-2022. Further investigation revealed that, during 1976-2010, areas that received above- normal rainfall during positive IOD experienced below-normal during 2011- 2022 and vice versa. The same pattern relationship was observed for negative IOD. Spatial analysis demonstrates that the percentage departure of rainfall across the region mirrors the standardized rainfall anomalies. The study highlights that the changing relationship between OND IOD and OND rainfall corresponds to the east-west shift of Walker circulation, as well as the north-south shift of Hadley circulation. Analysis of sea surface temperature (SST) indicates that both positive and negative IOD events strengthened during 2011-2022 compared to 1976-2010. Close monitoring of this relationship across different timescales could be useful for updating OND rainfall seasonal forecasts in Tanzania, serving as a tool for reducing socio-economic impacts.
文摘Understanding the characteristics of extreme rainfall is crucial for effective flood management planning, as it enables the incorporation of insights from past extreme rainfall patterns and their spatiotemporal distribution. This work investigated the changes in the frequency and pattern of extreme rainfall over Uganda, using daily datasets sourced from Climate Hazard Group InfraRed Precipitation with Station (CHIRPS-v2) for the period 1981 to 2022. The study utilized the extreme weather Indices provided by the Expert Team on Climate Change Detection and Indices (ETCCDI). Attention was directed towards September to November (SON) rainfall season with precise analysis of four indices (Rx1day, Rx5day, R95p, and R99p). The Sequential Mann-Kendall (SQMK) non-parametric test was applied to identify abrupt changes in SON extreme rainfall trends. Results showed that October consistently recorded the highest count of extreme rainfall days across all four indices. The long-term analysis revealed fluctuations in extreme rainfall events across years, with certain periods exhibiting heightened intensity. The analysis portrayed a shift in the decadal variations and region-specific distribution of extreme rainfall, with Eastern Uganda and areas around Lake Victoria standing out compared to other regions. The findings further revealed an increase in extreme rainfall for all indices in the recent decade (2011-2022) with 2019/2020 standing out as the extreme years of SON for the study period. While trendlines suggested a slight increase in intense daily rainfall events, the SQMK tests revealed statistical significance in the trend of prolonged periods of intense daily rainfall. This study contributes to the understanding of the spatiotemporal variability and trends of extreme rainfall events over Uganda during the SON season, which is crucial for the assessment of climate change impacts and adaptation strategies. It provides valuable information for seasonal extreme rainfall forecasting, development of early warning systems, flood risk management, and disaster preparedness plans.
基金primarily supported by the Chinese National Natural Science Foundation of China(Grant No. G42192553)Open Fund of Fujian Key Laboratory ofSevere Weather and Key Laboratory of Straits Severe Weather(Grant No. 2023KFKT03)+6 种基金the Open Project Fund of China Meteorological Administration Basin Heavy Rainfall Key Laboratory(Grant No. 2023BHR-Y20)the Open Fund of the State Key Laboratory of Remote Sensing Science (Grant No. OFSLRSS202321)the Program of Shanghai Academic/Technology Research Leader(Grant No. 21XD1404500)the Shanghai Typhoon Research Foundation (Grant No. TFJJ202107)the Chinese National Natural Science Foundation of China (Grant No. G41805016)the National Meteorological Center Foundation (Grant No. FY-APP-2021.0207)the High Performance Computing Center of Nanjing University of Information Science&Technology for their support of this work
文摘This paper presents an attempt at assimilating clear-sky FY-4A Advanced Geosynchronous Radiation Imager(AGRI)radiances from two water vapor channels for the prediction of three landfalling typhoon events over the West Pacific Ocean using the 3DVar data assimilation(DA)method along with the WRF model.A channel-sensitive cloud detection scheme based on the particle filter(PF)algorithm is developed and examined against a cloud detection scheme using the multivariate and minimum residual(MMR)algorithm and another traditional cloud mask–dependent cloud detection scheme.Results show that both channel-sensitive cloud detection schemes are effective,while the PF scheme is able to reserve more pixels than the MMR scheme for the same channel.In general,the added value of AGRI radiances is confirmed when comparing with the control experiment without AGRI radiances.Moreover,it is found that the analysis fields of the PF experiment are mostly improved in terms of better depicting the typhoon,including the temperature,moisture,and dynamical conditions.The typhoon track forecast skill is improved with AGRI radiance DA,which could be explained by better simulating the upper trough.The impact of assimilating AGRI radiances on typhoon intensity forecasts is small.On the other hand,improved rainfall forecasts from AGRI DA experiments are found along with reduced errors for both the thermodynamic and moisture fields,albeit the improvements are limited.
文摘为了准确监测和客观评估稻纵卷叶螟对水稻生长发育和产量形成的危害,利用ASD Field Spec3地物波谱仪和SPAD-502叶绿素仪分别采集控制大田试验(2015年和2019年)和自然大田试验(2020年)在各生育期(拔节期、孕穗期、灌浆期、成熟期)水稻的冠层高光谱数据和SPAD值,调查采集样点的虫量和水稻卷叶率,对比分析两种试验中稻纵卷叶螟的虫害发生特征、水稻冠层光谱特征和水稻生理生态参数特征,建立基于高光谱参数的水稻受稻纵卷叶螟危害的生理生态参数估算模型。结果表明,(1)两种试验的水稻SPAD值和冠层的红边至近红外波段的反射率均随着稻纵卷叶螟虫害程度的加重而降低,而可见光波段的反射率则相反;(2)自然大田试验的SPAD值和红光至近红外波段的冠层反射率在水稻生长发育前期要显著低于控制大田试验,而到了后期则反而要略高于控制大田试验;(3)综合分析筛选出自然大田试验和控制大田试验中的多个虫害特征参数和植被指数分别构建出了SPAD的单因子和多因子估算模型,各模型均达到了较好的估算效果,在单因子模型中EVI的二项式函数模拟效果最好,而多因子线性回归估测模型的模拟效果优于所有的单因子模型;(4)通过2021年对这些模型的应用检验发现:这些模型中基于虫量、卷叶率、OSAVI、EVI和DVI的单因子估算模型的SPAD估算值与实测值拟合度很高,其Rv 2均超过了0.8,达到了比较理想的估算效果,这为稻纵卷叶螟危害下的水稻SPAD值估测提供了一种精度较高且可行的估算方法。
