深度神经网络拾取地震P和S波到时

从地震波形数据中快速准确地提取各个震相的到时是地震学中的基础问题.本文针对上述问题提出了利用深度神经网络拾取到时的新方法，建立了用于地震到时提取的17层Inception深度网络模型，在对原始三分量数据进行高通滤波和归一化处理后输入网络直接输出到时信息.整个过程基于神经网络自适应提取波形特征，自动输出结果.通过对100组加了不同强度的噪声数据进行了可靠性检验，相比于其他方法神经网络方法对于噪声具有较高的容忍度以及稳定性，并且与地震目录数据有较高的相似性.相比于AR-AIC+STA/LTA，深度神经网络虽然运算速度稍慢，但整个过程不需设定时窗与阈值，同时具有更高的可用性，并且可以迭代升级以提高精度.此方法作为人工智能方法，为波形到时拾取提供了新思路.     

华北雨季开始早晚与大气环流和海表温度异常的关系

本文利用国家气候中心的1961~2016年华北雨季监测资料、美国国家环境预报中心/大气研究中心（NCEP/NCAR）的大气再分析资料、NOAA海表温度资料,分析了华北雨季开始早晚的气候特征,然后利用合成分析、回归分析等方法,研究了华北雨季开始早晚与大气环流系统和关键区域海表温度的关系。结果表明,56 a来华北雨季开始最早在7月6日,最晚在8月10日,1961~2016年华北雨季开始平均日期是7月18日。华北雨季开始时间具有显著的年际变化,但雨季发生早晚的长期变化趋势不太明显。华北雨季开始早晚与西太平洋副热带高压（简称副高）、东亚副热带西风急流、东亚夏季风等环流系统的活动关系密切,当对流层高层副热带西风急流建立偏早偏强,中层西太平洋副高第二次北跳偏早,低层东亚夏季风北进提前时,华北雨季开始偏早,反之华北雨季开始偏晚。华北雨季开始早晚与春、夏季热带印度洋、赤道中东太平洋海表温度关系显著且稳定,当Ni?o3.4指数和热带印度洋全区海表温度一致模态（IOBW）为正值时,贝加尔湖大陆高压偏强,副高偏强偏南,东亚夏季风偏弱,导致华北雨季开始偏晚;当海表温度指数为负值时,则华北雨季开始偏早。     

基于海温异常的南海夏季风爆发的可预报性分析

定义了综合评估南海地区大气对海温异常等外强迫响应的指数,并以此作为中间变量提出了利用前期海温异常预报南海夏季风爆发早晚的方法,进一步分析了基于海温异常的南海夏季风爆发的可预报性。分析表明:基于前期海温的异常,对于以日为单位预报对象的定量的南海夏季风爆发日期来说,基本没有可预报性;而对于定性的南海夏季风爆发早晚的预测,则可预报性大为提高。其预报时效可以提前至前期秋季的11月份。     

A New Scheme for Predicting Leaf Onset in Summer-Green Vegetation in the Northern Hemisphere

A modified thermal time model(MTM) was developed to reproduce the leaf onset for summer-green vegetation in the Northern Hemisphere. The model adopts the basic concept of a thermal time model(TM) in that leaf onset is primarily triggered by growing degree days(GDD). Based on global phenology data derived from satellite observations, a new parameterization for the critical model parameter Tb(i.e., baseline temperature for GDD calculation) has been introduced, and the spatial distribution of Tb was calculated. Simulations of leaf onset during 1982–2000 in the range 30–90°N showed a significant improvement of MTM over the standard TM model with constant Tb. The mean error and mean absolute error of the climatological simulation were 1.11 and 6.8 days, respectively, and 90% of the model error(5th and 95 th percentiles) was between-12.4 and 13.7 days.     

1990s年代际转型前后南海季风系统的季节变化

利用多变量经验正交分解(MV-EOF)等方法,研究了在季节变化尺度上南海季风系统的时空分布特征。结果表明:南海夏季风的爆发时间在1993—1994年前后存在显著的年代际转型,由爆发偏晚转变成爆发偏早。第一模态表现为冬夏反位相的年周期变化,但爆发早年夏季风持续时间略长于爆发晚年,空间上都反映了南海中央海盆区的夏季强降水和850 hPa上南海北部的气旋性环流异常,但夏季风爆发早年中国华南沿海降水加强而南海南部降水偏少。相应的大范围环流场上主要反映了南海夏季风爆发后进入盛夏时节亚太地区大范围的环流特征,南海夏季风爆发偏早年索马里越赤道气流偏强,东亚季风槽位置偏北,爆发偏晚年则相反。第二模态反映了南海季风系统春秋反位相的季节变化,且秋季的振幅更强,空间降水场上对应着秋季华南沿海和南海北部与南海中南部北旱南涝的跷跷板式分布,850 hPa风场上则主要表现为异常的东北季风,该模态时空特征表明南海夏季风爆发偏早年的秋季,冬季风建立也偏早,越南及周边地区的降水偏多。相应的大范围环流场上则主要反映了冬季风的环流特征,在南海夏季风爆发偏早年的秋季,菲律宾以东的热带对流减弱,PJ波列增强,爆发晚年则相反。     

Determination of onset date of the South China Sea summer monsoon in 2006 using large-scale circulations

Since the South China Sea (SCS) summer monsoon (SCSSM) is pronouncedly featured by abruptly intensified southwesterly and obviously increased precipitation over the SCS,the lower-tropospheric winds and/or convection intensities are widely used to determine the SCSSM onset.The methods can be used successfully in most of the years but not in 2006.Due to the intrusion of Typhoon Chanchu(0601)that year,the usual method of determining SCSSM onset date by utilizing the SCS regional indices is less capable of pinpointing the real onset date.In order to solve the problem,larger-scale situations have to be taken into account.Zonal and meridional circulations would be better to determine the break-out date of SCSSM in 2006.The result indicates that its onset date is May 16.Moreover,similar onset dates for other years can be obtained using various methods,implying that large-scale zonal and meridional circulations can be used as an alternative method for determining the SCSSM onset date.     

南海夏季风爆发与西北太平洋热带气旋活动

利用1965—2007年美国台风预警中心(JTWC)的热带气旋(TC)及NCEP/NCAR再分析资料,初步研究南海夏季风爆发与西北太平洋(包括南海)热带气旋活动之间的关系。结果表明,南海夏季风的爆发伴随着西北太平洋(尤其南海区域)TC生成个数和活动频数比爆发前有显著增加,而超过1/2的年份南海夏季风爆发前2候和爆发候西北太平洋(150°E以西)有TC活动,表明TC活动可能是南海夏季风爆发的触发机制之一。在大多数(77%)南海夏季风爆发偏早年,爆发前2候和爆发候西北太平洋TC活动偏多,且TC生成位置偏西;而大多数(77%)爆发偏晚年份,爆发前2候和爆发候没有TC活动。季风的偏早爆发受季节内振荡、西北太平洋TC活动、中纬度冷空气活动等复杂因素影响,而季风的偏晚爆发则主要受季节内振荡影响。     

Sub-seasonal Prediction of the South China Sea Summer Monsoon Onset in the NCEP Climate Forecast System Version 2

This study depicts the sub-seasonal prediction of the South China Sea summer monsoon onset (SCSSMO) and investigates the associated oceanic and atmospheric processes, utilizing the hindcasts of the National Centers for Environmental Prediction (NCEP) Climate Forecast System version 2 (CFSv2). Typically, the SCSSMO is accompanied by an eastward retreat of the western North Pacific subtropical high (WNPSH), development of the cross-equatorial flow, and an increase in the east-west sea surface temperature (SST) gradient. These features are favorable for the onset of westerlies and strengthening of convection and precipitation over the South China Sea (SCS). A more vigorous SCSSMO process shows a higher predictability, and vice versa. The NCEP CFSv2 can successfully predict the onset date and evolution of the monsoon about 4 pentads (20 days) in advance (within 1–2 pentads) for more forceful (less vigorous) SCSSMO processes. On the other hand, the climatological SCSSMO that occurs around the 27th pentad can be accurately predicted in one pentad, and the predicted SCSSMO occurs 1–2 pentads earlier than the observed with a weaker intensity at longer leadtimes. Warm SST biases appear over the western equatorial Pacific preceding the SCSSMO. These biases induce a weaker-than-observed WNPSH as a Gill-type response, leading to weakened low-level easterlies over the SCS and hence an earlier and less vigorous SCSSMO. In addition, after the SCSSMO, remarkable warm biases over the eastern Indian Ocean and the SCS and cold biases over the WNP induce weaker-than-observed westerlies over the SCS, thus also contributing to the less vigorous SCSSMO.     

1998年夏季季风爆发前后南海上层环流的诊断分析

基于1998年南海季风试验（SCSMEX，South China Sea Monsoon Experiment)期间2个强化观测航次（4－5月及6－7）所获温盐深（CTD）资料，利用一个改进逆模式研究了夏季季风爆发前后南海环流的演变特征。诊断计算表明，在此期间南海环流主要表现为两脊两槽型，即越南以东和菲律宾以西呈反气旋式环流，南海北部和南海中部呈气旋式环流。但对局部区域而言，可以发现在季风爆发前后其环流结构有明显的差异。上述计算结果亦与等压面上海水密度分布的定性分析结果及同期观测的ADCP资料进行了比较，结果表明模式计算所得到的南海上层环流主要特征与定性分析结果及实测资料大体一致，诊断结果可作为南海上层季风环流演变机制研究的依据。