Winds that flow within a forest canopy in complex terrain exhibit considerable variability that can have significant consequences for environmental processes such as the spread of bushfires, seed dispersal and transport of pollutants. This variability is still poorly understood, due in part to limited observations of canopy winds over complex terrain, and as such is often unaccounted for in deterministic modelling approaches. Probabilistic representation of wind fields can better characterise variability and build understanding of uncertainty in applications such as bushfire prediction. This study introduces two new, publicly available, datasets that support analyses of canopy winds over complex terrain using novel statistical approaches. Wind fields are characterised using bivariate distributions, representing the response of synoptic winds to changing landscapes, including post-bushfire forest regrowth and varying topography across uniform vegetation. For statistical comparison, non-parametric Kolmogorov-Smirnov style tests are considered. A new test is proposed for the comparison of bivariate circular distributions arising from analyses of wind direction. The study broadly reaffirms established theory while highlighting the variability of canopy winds over complex terrain and the importance of developing statistical techniques that result in a better quantitative understanding of these wind fields and the processes driving them.

在复杂地形中的林冠层中流动的风表现出很大的变异性，会对环境过程产生重大影响，例如丛林大火的蔓延，种子的扩散和污染物的运输。由于对复杂地形上的冠层风的观测有限，因此对这种可变性的了解仍然很少，因此，在确定性建模方法中通常无法解决这一问题。风场的概率表示可以更好地表征变化性，并增强对诸如林区大火预报等应用程序中不确定性的理解。这项研究引入了两个新的公开可用的数据集，这些数据集支持使用新颖的统计方法分析复杂地形上的冠层风。使用双变量分布来表征风场，代表天气对变化的景观的响应，包括篝火晚会后的森林再生和均匀植被的地形变化。为了进行统计比较，考虑使用非参数Kolmogorov-Smirnov样式检验。为了比较风向分析产生的二元圆分布，提出了一种新的检验方法。这项研究在很大程度上重申了既定的理论，同时强调了复杂地形上冠层风的变化性，以及发展统计技术的重要性，以使人们对这些风场及其驱动过程有更好的定量理解。为了比较风向分析产生的二元圆分布，提出了一种新的检验方法。这项研究在很大程度上重申了既定的理论，同时强调了复杂地形上冠层风的变化性，以及发展统计技术的重要性，以使人们对这些风场及其驱动过程有更好的定量理解。为了比较风向分析产生的二元圆分布，提出了一种新的检验方法。这项研究在很大程度上重申了既定的理论，同时强调了复杂地形上冠层风的变化性，以及发展统计技术的重要性，以使人们对这些风场及其驱动过程有更好的定量理解。