calculating linear trend

Sachin Ghude sachinghude at TROPMET.RES.IN
Tue Apr 3 05:55:18 EDT 2007


Dear all,

I am trying to Calculate linear trend defined by least square fitting from 10 years of
monthly data. For calculating linear trend I have used following method and obtained
the results. ( I have taken help from Bin Guan trend.gs script placed at his website)

However, I am not sure whether above method is correct or I am doing any mistake. Would
somebody please advice me if I am doing any mistake while using following method.

ga->open G:Sno2\start.ctl
Scanning description file:  G:sno2\start.ctl
Data file G:\SNO2\%m2%y4.dat is open as file 1
LON set to 0 360
LAT set to -89.875 89.875
LEV set to 1 1
Time values set: 1996:1:1:0 1996:1:1:0
-----------------------------------------------------------
Then I obtained dimension information as

ga->query dims
Default file number is: 1
X is varying   Lon = 0 to 360   X = 720.5 to 2160.5
Y is varying   Lat = -89.875 to 89.875   Y = 1 to 720
Z is fixed     Lev = 1  Z = 1
T is fixed     Time = 00Z01JAN1996  T = 1

I used x and y values for linear trend calculation
--------------------------------------------------------------
set t 1 132 (10 years monthly data)
define ones=1
define nno2=tloop(sum(ones,t=1,t+0))+1-1
define nno2=maskout(nno2,maskout(no2,no2)/maskout(no2,no2))
define nno2=ave(ave(nno2,x=720.5,x=2160.5),y=1,y=720)
set t 1
define trendslope=tregr(nno2,maskout(no2,no2),t=1,t=132)
define a=trendslope
set gxout shaded
display a

after running above commands I could obtained the result. However, I am not sure
whether above method is correct or I am doing any mistake.

with regards

Sachin Ghude



More information about the gradsusr mailing list