去年研究了波折射指数的计算公式,当时觉得太复杂了不想写代码于是放弃,但今年突然又觉得还是算一下,于是又研究了几天,又被折磨了一下。人果然不能偷懒!
利用波折射指数诊断波的传播环境,计算方法:(Andrews et al.,1987)
公式都差不多,但Weinberger et al.,(2021)这篇文章里面改变了好几个形式,我就算了一个比较原始的。
波折射指数体现大气环流的背景场是否有利于行星波传播;但其计算是基于纬圈平均的场,实际波的反射传播是一个局地的区域现象,波折射指数只能判断背景流是否有利于波动反射,但无法准确地表征波动反射的发生,当RI<0时,行星波发生反射;RI>0,行星波在此区域传播,越大越容易传播;(注意是波折射指数的平方RI,看文献好像没有找到识别局地区域反射的指数,还看见一个U2-U10的定义,感觉是风场的切变,定义方法比较粗糙)。
Albers and Birner., (2014)文章中RI的单位是m-2,根据(f/2NH)2,f单位为s-1,NN单位为s-2,H单位为m,倒推第三项单位为m-2,合理;计算结果乘了地球半径的平方,将RI无量纲化;感觉公式算出来量级有一点奇怪,算出来主要还是以位涡第一项为主导的样子;
(All fields are multiplied by Earth’s radius squared, which nondimensionalizes the refractive index and gives the PV gradient the units of meters per second)
浅浅的用ncl算了一下,结果感觉看起来还算正常?但是第一项量级有一点问题;真烦人啊啊啊啊啊啊啊啊啊啊看了一下文献中的量级差不多(Li et al.,2017)
Li et al.,2017文章中说正负折射指数的平方会相互抵消,新定义一个负波折射指数平方的频率Fn2来诊断行星波的传播,具体的计算方法,将格点上n2为负值的一天定义为1,为正值的一天定义为0,Fn2定义为负值的天数除以总天数,Fn2合成的结果减去气候态的得到异常值,行星波在Fn2小的区域传播,在大的区域反射;
···
;============================================================
; calculate QG RI index(Andrews et al.,1987)
;============================================================
;pv:potential vorticity;
;T :basic state of air temperature;
;u :basic state of uwind;
;k :zonal wavenumber(k=1 means wave 1)
;Two methods:By using the pv data and calculting the pv;
;if the phase speed c equals zero, the RI is treated as the stationary wave RI;
;return:RI Uyy Uzz RI1
undef("def_cal_QGRI")
function def_cal_QGRI(pv[][][][]:numeric,T[][][][]:numeric,u[][][][]:numeric,k[*]:numeric)
local sclhgt,gc,pi,dName,level,lat,lon,re,f,density,coslat,leveltmp,coslattmp,ftmp,dthetadz,NN,pv_ZonalMean,u_ZonalMean,T_ZonalMean,k,T_ZonalMean,densitytmp,u_grad,uyy1,u_gradz,uzz1,pv_grad2,RI2,RI3
begin
;------ scale height
sclhgt = 7000.;units:m
; Gas constant
gc = 290.
pi = atan(1.0)*4.
; 自转角速度
wmg=2.*pi/(60.*60.*24.)
dName = getvardims(pv)
level = pv&$dName(1)$
lat = pv&$dName(2)$
lon = pv&$dName(3)$
dim =dimsizes(T)
;------Radius of the earth
re = 6.37*10^6 ;6378388 units:m
;------Coriolis parameter
f = lat(:)
f = (/2.*2.*pi/(60.*60.*24.)*sin(pi/180.*f)/)
;------ cosine
coslat = cos(lat(:)*pi/180.)
coslat@_FillValue = default_fillvalue(typeof(coslat))
coslat = (/where(coslat.le.0.,coslat@_FillValue,coslat)/)
;zanal ano
T_ZonalMean=dim_avg_n_Wrap(T, 3)
printVarSummary(T_ZonalMean)
; 1-D -> 3-D
leveltmp = conform_dims(dimsizes(T_ZonalMean),level,1)
coslattmp = conform_dims(dimsizes(T_ZonalMean),coslat,2)
ftmp = conform_dims(dimsizes(T_ZonalMean),f,2)
;densitytmp = conform_dims(dimsizes(T_ZonalMean),density,1)
;ρ0=ρs*exp(-z/H)
densitytmp=1.2*(leveltmp/1000)
printVarSummary(leveltmp)
printVarSummary(coslattmp)
printVarSummary(ftmp)
printVarSummary(densitytmp)
;-------------------------------------------------------------------------------
; vertical gradient of potential temperature (K/m)
dthetadz = center_finite_diff_n(T_ZonalMean*(1000./leveltmp)^0.286,-sclhgt*log(level/1000.),False,0,1)
printVarSummary(dthetadz)
;--------------------------------------------------------------------
; Brunt Vaisala frequency
NN = (gc*(leveltmp/1000.)^0.286)/sclhgt * dthetadz
NN@var_desc = "Brunt Vaisala frequency"
NN@units = "1/s^2"
NN@long_name = "basic state Brunt Vaisala frequency (TN2001)"
NN@_FillValue = T@_FillValue
NN = where(NN.gt.0,NN,NN@_FillValue)
printVarSummary(NN)
print("-----NN-----")
printMinMax(NN, 0)
;------ zonal-mean zonal wind
u_ZonalMean=dim_avg_n_Wrap(u, 3)
printVarSummary(u_ZonalMean)
print("-----u_ZonalMean-----")
printMinMax(u_ZonalMean, 1)
;------------------------------------------------------------------------------------------------
;---------------------------求zonal mean的pv经向梯度 pv_grad
;第1项:2*自转角速度/地球半径*cos纬度
pv_grad1 =2.*wmg/re*coslattmp
print("------pv_grad1-----")
printMinMax(pv_grad1, 0)
;注意没有加负号!-uyy -uzz
;-----第2项:uyy
u_grad =center_finite_diff_n(u_ZonalMean*coslattmp ,lat*pi/180.,False,0,2)
uyy1 =center_finite_diff_n(u_grad/coslattmp ,lat*pi/180.,False,0,2)
uyy =1./(re^2)*uyy1
print("-----uyy-----")
printMinMax(uyy, 0)
;-----第3项:uzz
u_gradz =center_finite_diff_n(u_ZonalMean,-sclhgt*log(level/1000.),False,0,1)
printVarSummary(u_gradz)
uzz1 =center_finite_diff_n(u_gradz*densitytmp/NN,-sclhgt*log(level/1000.),False,0,1)
printVarSummary(uzz1)
uzz=ftmp^2/densitytmp*(uzz1)
print("-----uzz-----")
printMinMax(uzz, 0)
print("-----cal pv_grad-----")
pv_grad=(/pv_grad1-uyy-uzz/)
printMinMax(pv_grad, 0)
copy_VarCoords(u_ZonalMean, pv_grad)
printVarSummary(pv_grad)
;=======================================================================
;直接使用pv的数据计算
;--------------------------------------------
pv_grad2=grad_latlon_cfd(pv, lat*pi/180., lon*pi/180., True, False)
;0为经向梯度
printVarSummary(pv_grad2)
;------ 对lon维 zonal-mean potential vorticity
pv_ZonalMean=dim_avg_n_Wrap(pv_grad2[0], 3)
print("-----data pv_grad-----")
printMinMax(pv_ZonalMean, 0)
printVarSummary(pv_ZonalMean)
;=======================================================================
;------QG RI index for the stationary wave k(c=0)
print("-----RI1-----")
RI1=pv_grad/u_ZonalMean ;有一点问题感觉!为什么量级差异这么多啊啊啊啊啊啊啊啊!!!!!!10^-10/10^1 量级不对哇啊啊啊啊!
printMinMax(RI1, 0) ;
copy_VarCoords(u_ZonalMean, RI1)
print("-----RI2-----")
RI2=(k/(re*coslattmp))^2
printMinMax(RI2, 0)
print("-----RI3-----")
RI3=(ftmp/(2.*sclhgt))^2/NN
printMinMax(RI3, 0)
print("-----RI-----")
RI=(/RI1-RI2-RI3/)
printVarSummary(RI)
printMinMax(RI, 0)
RI=RI*re^2 ;*地球半径的平方,将折射率无量纲化(Albers and Birner,2014);
;根据(f/2NH)^2 f单位为s-1,NN单位为s-2,H单位为m,倒推第三项单位为m-2,合理!
copy_VarCoords(T_ZonalMean, RI)
RI@units = "1"
RI@long_name = "QGRI"
;--------------------不知道返回uyy和uzz时需不需要这样处理,输出看一下结果!
uyy=-uyy*re^2
uzz=-uzz*re^2
pv_grad=pv_grad*re^2
RI1=RI1*re^2
copy_VarCoords(T_ZonalMean, uyy)
copy_VarCoords(T_ZonalMean, uzz)
uyy@long_name = "uyy"
uzz@long_name = "uzz"
copy_VarCoords(T_ZonalMean, pv_grad)
pv_grad@long_name = "pv_grad"
pv_grad@units = "m/s"
copy_VarCoords(T_ZonalMean, RI1)
RI1@long_name = "RI1"
print("-----re2-----")
printMinMax(RI, 0)
return([/RI,uyy,uzz,pv_grad,RI1/])
end
代码包括直接用pv数据计算梯度和用公式计算的,改一下输入就行,输出的变量需要啥就输出啥!
参考文献:
Andrews, D. G., J. R. Holton, and C. B. Leovy, 1987: Middle Atmosphere Dynamics. International Geophysics Series, Vol. 40, Academic Press, 489 pp.
Albers, J.R., Birner, T., 2014. Vortex Preconditioning due to Planetary and Gravity Waves prior to Sudden Stratospheric Warmings. Journal of the Atmospheric Sciences 71, 4028–4054.. https://doi.org/10.1175/jas-d-14-0026.1
Li, Q., Graf, H.-F., & Giorgetta, M. A. (2007). Stationary planetary wave propagation in Northern Hemisphere winter-climatological analysis of the refractive index. Atmospheric Chemistry and Physics, 7(1), 183–200. https://doi.org/10.5194/acp-7-183-2007
Weinberger, I., Garfinkel, C.I., White, I.P., Birner, T., 2021. The Efficiency of Upward Wave Propagation near the Tropopause: Importance of the Form of the Refractive Index. Journal of the Atmospheric Sciences 78, 2605–2617.. https://doi.org/10.1175/jas-d-20-0267.1