西北太平洋是全球台风发生频率最高、强度最大的海域。我国位于西北太平洋西北部，几乎每年都会受到台风灾害影响并造成严重损失。气候变化背景下，系统性灾害异常现象也趋于频繁，如北上台风的频率增加，以及内陆城市（如郑州）受台风灾害影响的潜在威胁趋于严重等，而这些现象对于基于传统方法的台风概率模拟及其风险评估的框架来说难以解释，也具有关键性的技术挑战。因此，建立台风概率与系统性气候因子的对应关系将对台风灾害风险评估和防治具有重要意义。

由于副热带高压影响范围大、系统性强，且与台风形成与运动相关，因此本文主要选取副热带高压为主要气候因子，基于历史台风数据与副热带高压数据，首先分析了历史台风致灾因子与副热带高压的时空分布特征；其次，基于西北太平洋区域台风生成、运动机理，量化分析了历史台风路径与副热带高压的空间相关关系，并据此构建了台风路径概率模拟模型；然后，基于历史风场数据，统计分析了不同强度等级风速空间分布概率，结合西北太平洋台风路径概率分布得到不同强度台风概率分布；最后，依据台风路径概率模型随机生成事件级台风路径，构建了巨灾情景下目标区域的台风危险性空间分布模拟方法。本文主要结果如下：

（1）历史台风年发生频次有所下降，强台风（STY）与超强台风（SuperTY）的频次有所增加；空间上，历史台风路径随经纬度分布变化符合二项高斯分布，随研究区经度自西向东移、纬度自南向北移，台风路径概率先增大后减小，台风路径高频分布区域多集中在10°N-30°N，120°E-150°E范围内，且台风路径点有北移趋势。

（2）副热带高压的面积与强度指数在1978年-1996年间整体呈波动下降趋势，1996年-2020年间整体呈波动上升趋势，且最大值出现在2020年，空间上，随着时间的推移，西北太平洋副热带高压频次与范围有先收缩再扩张增强的趋势；21世纪初与20世纪末相比副热带高压频次整体增大，东部及北部区域尤为明显，而在我国台湾东南部频次有所减小。

（3）历史台风路径与副热带高压边缘相关关系分析结果表明，二者具有良好的空间相关关系，其平均相关性系数为0.49；针对台风路径与副热带高压边缘之间的距离变化的空间关系分析，二者关系符合二次函数分布，即随经度自西向东、纬度自南向北，台风路径与副热带高压边缘之间的距离呈现先减小后增大趋势。

（4）构建了基于副热带高压的台风路径概率模型，基于该模型模拟的台风路径分布概率特征与历史台风路径分布特征基本一致；与传统模拟方法相比，该结果拥有更大分布范围的台风路径概率值，其中，我国内陆、甚至中高纬度地区均给出了台风概率。本文不仅从宏观角度提供了全域台风路径概率计算的方法和分布结果，还可以从地区防灾减灾需求角度，构建指定目标区域的随机事件级台风情景，对一定强度下的台风巨灾事件场景进行科学模拟，从而实现科学防灾减灾。

The Northwest Pacific is the sea area with the highest frequency and intensity of tropical cyclones worldwide. China is located in the northwest of the Northwest Pacific and is affected by tropical cyclone disasters almost every year, causing serious losses. In the context of climate change, systematic disaster anomalies are also becoming more frequent, such as an increase in the frequency of northward tropical cyclones, and the potential threat of inland cities (such as Zhengzhou) being affected by tropical cyclone disasters becoming more severe. These phenomena are difficult to explain and pose key technical challenges for the framework of probability simulation and risk assessment of tropical cyclones based on traditional methods. Therefore, establishing the corresponding relationship between the probability of tropical cyclones and systemic climate factors will be of great significance for the risk assessment and prevention of tropical cyclone disasters.

Due to the large impact range and strong systematicity of subtropical high, as well as its correlation with the formation and movement of tropical cyclones, this article mainly selects subtropical high as the main climate factor. Based on historical tropical cyclone data and subtropical high data, the spatiotemporal distribution characteristics of historical tropical cyclone disaster factors and subtropical high are first analyzed; Secondly, based on the generation and motion mechanisms of tropical cyclones in the Northwest Pacific region, the spatial correlation between historical tropical cyclone tracks and subtropical high was quantitatively analyzed, and a probability simulation model for tropical cyclone tracks was constructed based on this; Then, based on historical wind field data, the spatial distribution probability of wind speeds at different intensity levels was statistically analyzed, and the probability distribution of tropical cyclones with different intensities was obtained by combining it with the probability distribution of tropical cyclone tracks in the Northwest Pacific; Finally, based on the probability model of tropical cyclone tracks, event level tropical cyclone tracks were randomly generated, and a simulation method for the spatial distribution of tropical cyclone hazards in the target area under catastrophic scenarios was constructed. The main conclusions of this article are as follows:

(1) The annual frequency of historical tropical cyclones has decreased, and the frequency of strong typhoons (STY) and super typhoon has increased; In space, the change of historical tropical cyclone track with longitude and latitude Degree distribution conforms to binomial Gaussian distribution. With the longitude of the study area moving from west to east and latitude moving from south to north, the probability of tropical cyclone track first increases and then decreases. The high-frequency distribution area of tropical cyclone track is mostly concentrated in the range of 10 ° N-30 ° N, 120 ° E-150 ° E, and the tropical cyclone track point has a trend of moving northward.

(2) The area and intensity index of the subtropical high showed a fluctuating downward trend from 1978 to 1996, and an overall fluctuating upward trend from 1996 to 2020, with the maximum value appearing in 2020. spatially, with the passage of time, the frequency and range of the subtropical high in the northwest Pacific showed a trend of first contracting, then expanding and strengthening; Compared with the end of the 20th century, the frequency of the subtropical high increased overall in the early 21st century, particularly in the eastern and northern regions, while the frequency decreased in the southeastern part of Taiwan, China.

(3) There is a good spatial correlation between the tropical cyclone track and the subtropical high, the average spatial correlation between the two is 0.49. For the distance between the tropical cyclone track and the edge of the subtropical high, the spatial relationship between the two conforms to the quadratic function, that is, it moves from west to east with the longitude of the study area, and from south to north with the latitude, the distance between the tropical cyclone track and the edge of the subtropical high first decreases and then increases.

(4) A probability model of tropical cyclone tracks based on subtropical high has been constructed, and the probability characteristics of simulated tropical cyclone tracks based on this model are basically consistent with the distribution characteristics of historical tropical cyclone tracks; Compared with traditional simulation methods, this result has a larger distribution range of probability values for tropical cyclone tracks, with tropical cyclone probabilities given in inland and even mid to high latitude regions of China. This article not only provides a method and distribution results for calculating the probability of global tropical cyclone tracks from a macro perspective, but also constructs a random event level tropical cyclone scenario in a designated target area from the perspective of regional disaster prevention and reduction needs. It scientifically simulates the tropical cyclone catastrophic event scenario at a certain intensity, thereby achieving scientific disaster prevention and reduction.