若是逐日资料，则原始逐日资料减去对应这一日的气候态，就去掉了年循环（或者季节循环），只剩下低频信号，年际信号，年代际信号了。

若原始资料是逐月资料，则减去对应月份的气候态，就去掉了年循环，只剩下年际、年代际等异常信号了。

去除全球变暖趋势，可采用直接去掉最小二乘线性趋势

Hi, Jianping

With regarding to your question about figure 4 you sent me by Wechat, here is my explanation. If you horizontal wind to indicate circulation change, there is no need to remove global mean.

      When you  use geopotential height as an indicator of anomalous circulation you need to remove the global mean since global warming will raise geopotential height everywhere.  Anomalous circulation is related to the gradient of geopotential height and therefore global mean height does not contribute to circulation change and removing global mean height does not affect circulation at all. Removing zonal mean further emphasize wave like pattern of circulation since it only remove a constant of zonal wind at each latitude circle, but does not affect meridional wind at all.  Therefore,  in order to illustrate anomalous circulation changes with global warming by using height variable, you have to remove the global mean. In order to emphasize zonally asymmetric of circulation (wave pattern), you can remove zonal mean as well. Hope this is clear.

Regards

Buwen  

计算说明：1. 1979-2013年夏季（JJA平均）ERAI的海平面气压和500 hPa位势高度分别减去各自的全球平均，再去趋势，最后与基于ERAI计算的已去趋势的HWI的相关关系，以及N216中historical的海平面气压和500 hPa位势高度分别减去各自的全球平均，再去趋势，最后与基于historical计算的去趋势的热浪指数的相关系数空间分布，其中相关系数计算是利用NCL自带的皮尔逊相关的函数计算得来。

计算说明：1. 1979-2013年夏季（JJA平均）ERAI的海平面气压和500 hPa位势高度分别减去各自的纬向平均，再去趋势，最后与基于ERAI计算的已去趋势的HWI的相关关系，以及N216中historical的海平面气压和500 hPa位势高度分别减去各自的纬向平均，再去趋势，最后与基于historical计算的去趋势的热浪指数的相关系数空间分布，其中相关系数计算是利用NCL自带的皮尔逊相关的函数计算得来。