Utilizing conventional meteorological observation data from automatic weather stations, a method based on Gradient Boosting Regression Tree model is proposed to correct the SSRD values derived from the ERA5 reanalysis dataset, produced by the European Centre for Medium-Range Weather Forecasts. Data from Guangzhou Station and Shantou Station from January 1, 2019 to December 31, 2021 were utilized for modeling training and testing. A cross-validation experiment was conducted to evaluate the transferability of the correction model of SSRD. The correlation coefficient between the monitoring estimate and the observation increased from 0.801-0.830 to 0.829-0.881, and the root mean square error decreased from 169.8-171.7 W/m2 to 155.5-159.8 W/m2. Further analysis shows that, in addition to the corrected element ERA5 SSRD, the solar altitude angle is the most important input in the Gradient Boosting Regression Tree model of total solar irradiance. Other influential variables, ranked by importance, include relative humidity, total cloud cover, near surface temperature and visibility.
近年来,社会对能源的需求越来越迫切,不可再生资源正逐渐被耗尽。太阳能作为永久的可再生能源,是未来发展必然选择的持续能源和战略能源,太阳能的分布和变化状况,受到越来越多的关注[4-5]。太阳辐射的时空分布及其变化规律、气候条件密切影响着光热利用系统的分布[6]。太阳辐射中功率占比最大的部分为可见光波段,太阳辐射覆盖了晶体硅主要吸收的300~1 127 nm波段[7],地表的太阳向下总辐照度(Surface Solar Radiation Downwards, SSRD)是光伏发电效率最重要的直接天气影响因素[8]。为了缓解能源危机,以光伏发电为代表的可再生能源发展迅速,但目前光伏发电量的预测精度较低,准确预测光伏发电功率对于电力系统的安全运行和经济运营至关重要[9],SSRD是光伏发电量最重要的影响因子,因此成为重要的大气监测对象。
我国地面自动气象观测站监测的主要物理要素包括气压、气温、湿度、风向、风速、降水、地表温度、云量、蒸发等,能够监测辐照度的地面自动气象观测站较少。地面观测站点的太阳总辐照度是与SSRD定义最为接近的物理量,两者都是太阳短波辐射(可见光和紫外波段)到达地面的辐照度,是直接辐射与散射辐射的总和,名称中的“向下”主要是为了与地球射向天空的长波辐射进行区分。对SSRD的监测主要是通过气象卫星所收集的遥感数据结合大气辐射传输模式完成,例如美国国家航空航天局下属云与地球辐射系统(The Clouds and the Earth’s Radiant Energy System, CERES)发布的SYN1deg反演产品[10]。利用资料同化方法和数值模式得到的再分析资料,也可形成全球网格化SSRD监测资料,如欧洲中期天气预报中心(ECMWF)的ERA5再分析资料就包含0.1°×0.1°水平分辨率的SSRD监测数据[11]。由于臭氧、水汽等物理要素垂直廓线缺少高频率的直接观测数据[12],通过辐射传输模式或资料同化方法对SSRD的估计与地基直接测量之间存在一定偏差。目前,我国安装有辐射观测设备的地面自动气象站分布较为稀疏,如果能利用地面观测的常规气象要素对辐射传输模式或资料同化方法生成的SSRD监测产品进行订正,将有利于提高SSRD的区域监测能力。
LIXiaowen, LIWeiliang, ZHOUXiuji.Analysis of the solar radiation variation of China in recent 30 years[J].Journal of Applied Meteorological Science,1998,9(1):24-31.(in Chinese)
[3]
肖子牛.太阳活动对地球气候的影响[J].自然杂志,2021,43(6):408-419.
[4]
XIAOZiniu.Impact of solar activity on Earth’s climate[J].Chinese Journal of Nature,2021,43(6):408-419.(in Chinese)
YANJiahai, ZHANGDongfeng, ANWei,et al.Distribution characteristics and assessment of solar energy resource in Shanxi[J].Journal of Arid Meteorology,2014,32(5):712-718.(in Chinese)
YAOYubo, ZHANGLiping, LIUWenzhu,et al.Research on application of different back reflection structures in silicon heterojunction solar cells[J].Acta Energiae Solaris Sinica,2022,43(10):37-42.(in Chinese)
ZHANGXi, KANGChongqing, ZHANGNing,et al.Analysis of Mid/Long term random characteristics of photovoltaic power generation[J].Automation of Electric Power Systems,2014,38(6):6-13.(in Chinese)
SHIHaoqi, GUOLi, LIUYixin,et al.Short-term forecasting of photovoltaic power based on total irradiance correction of multi-source meteorological forecast[J].Electric Power Automation Equipment,2022,42(3):104-112.(in Chinese)
[19]
DoellingD R, SunMoguo, NguyenL T,et al.Advances in geostationary-derived longwave fluxes for the CERES synoptic (SYN1deg) product[J].Journal of Atmospheric and Oceanic Technology,2016,33(3):503-521.
[20]
CamargoL R, SchmidtJ.Simulation of multi-annual time series of solar photovoltaic power:is the ERA5-land reanalysis the next big step?[J].Sustainable Energy Technologies and Assessments,2020,42:100829.
[21]
CARLEERM R, BOONEC D, WALKERK A,et al.Validation of water vapour profiles from the Atmospheric Chemistry Experiment (ACE)[J].Atmospheric Chemistry and Physics Discussions,2008,8(2):4499-4559.
[22]
JohnsonN E, BonczakB, KontokostaC E.Using a gradient boosting model to improve the performance of low-cost aerosol monitors in a dense,heterogeneous urban environment[J].Atmospheric environment,2018,184:9-16.
[23]
CHENH G, ZHANGH, BoningD,et al.Robust decision trees against adversarial examples[C]//Proceedings of the 36th International Conference on Machine Learning.Long Beach:PMLR,2019:1122-1131.
[24]
全国气象防灾减灾标准化技术委员会. 紫外线指数预报方法:GB/T 36744-2018 [S].北京:中国标准出版社,2018.National Standardization Technical Committee for Meteorological Disaster Prevention and Reduction.Forecasting method for ultraviolet index:GB/T 36744-2018 [S].Beijing:China Standard Press,2018.(in Chinese)
SIYang, ZHANGHaifeng.Prediction method of solar irradiance based on neural network[J].Journal of Qinghai University (Natural Science),2013,31(1):14-18.(in Chinese)
CHONGWei, SHAYizhuo, XINGHongyan,et al.A new correction algorithm for diffuse irradiance measured with rotating shadow-band pyranometer based on support vector regression[J].Acta Optica Sinica,2012,32(1):0112001.(in Chinese)
JIANGJunxia, GAOXiaoqing, QingquanLYU,et al.Study on cloud tracking and solar irradiance ultra-short-term forecasting based on TSI images[J].Acta Energiae Solaris Sinica,2020,41(5):351-358.(in Chinese)
WUHuiqin, SHIMin, ZHAOZengbao,et al.Stepwise revision of solar irradiance based on EC fine-grid numerical forecast products[J].Meteorological Science and Technology,2020,48(5):752-757.(in Chinese)
ZHANGH L, YUANY B, XUY J,et al.Remote sensing method for detecting green tide using HJ-CCD top-of-atmosphere reflectance[J].International Journal of Applied Earth Observation and Geoinformation,2021,102:102371.
[35]
HersbachH, BellB, BerrisfordP,et al.The ERA5 global reanalysis[J].Quarterly Journal of the Royal Meteorological Society,2020,146(730):1999-2049.