Cloud-top Relief Spatial Displacement Adjustments of GOES-R Images
description
Transcript of Cloud-top Relief Spatial Displacement Adjustments of GOES-R Images
Cloud-top Relief Spatial Displacement Adjustments
of GOES-R ImagesShayesteh Mahani
CREST & CE Dept. at the City College (CCNY) of the City University of New York (CUNY)
Co-Is: Johnny Luo, William Rossow, Reza Khanbilvardi, and Kibre Tesfagiorgis, CREST, CCNY/CUNY
NOAA-Collaborators: Robert Rabin, NOAA/NSSL, UW-Madison/CIMSS
Andrew Heidinger and Robert Kuligowski, NESDIS-STAR,
NOAA Satellite Science Week, April 30 – May 04, 2012, Kansas City
Objectives• Estimate Cloud-Top Height (CTH) using 3-D principals and vertical
height-temperature profile;
• Enhance the quality of high resolution GOES-R infrared (IR) data by adjusting pixel based cloud-top relief Spatial Displacements (SD).
Introduction Cloud VIS and IR images posses SD associated with CTH that is a
function of satellite height, satellite-cloud view angle, sun zenith angle, and earth curvature.
Stereo-graphical (3-D) principle using corresponding & simultaneous cloud IR data from GOES-E and -W, are applied to estimate CTH and its associated SD for adjustment.
The 3-D based technique and the GOES-R Algorithm Theoretical Basis Document (ATBD) will be evaluated against CTH from CALIPSO as a reference to examine if the combined method can improve CTH output.
Major Tasks & Accomplishments1. Derive the geometric relationship between cloud-top height and
pixel-based spatial displacement for GOES-R (accomplished);
2. Derive the relationship between CTH, X-parallax of SD, and IR-CTBT, utilizing 3-D principals from scan-synchronous GOES-E and -W IR Images and optimize the parameters of the relationship using the SCE (Shuffled Complex Evolution) calibration Model to estimate CTH and adjust SD (90% accomplished);
3. Understand which IR channels, and/or combination(s) of which IR channels work better to estimate CTH and to adjust CTSD for a given cloud type, region, and season (Year-2);
4. Evaluate the CTH estimates and the CTH from GOES-R ATBD algorithm against the CALIPSO-CTH product (Year-2);
Major Tasks & Future Work5. Modify/update the number and value of the optimized model
parameters of the CTH-CTBT or CTBT-SD relationships for seasonal, topography, land type, and storm type variability, (Years-2 & 3);
6. Investigate improvement of CTH estimates by incorporating the proposed technique into ATBD model and evaluating the new CTH product against CALIPSO-CTH (Year-3);
OUTLINE • Geometry of the relationship between CTH and pixel-based SD
• Relationship between CTH and CTBT and CTH-based X-parallax,
• Deriving the relationship between CTH, SD and IR-CTBT utilizing corresponding pairs of scan-synchronous and simultaneous GOES-E (13) and -W (11) as proxy for GOES-R Images.
• Validation & Results.
BUSSTOP 200
Satellite
Spatial Displacement related to Cloud-Top Height
= (h, , )
Geometric Relationship between Cloud Top Height and its related SD. Cloud-Top Relief Spatial Displacement (BC or ) depends on Height of Cloud (h), Distance from Satellite () and Satellite Height (H).
Cloud-Top Height-Displacement Relationship
= Satellite View Angle
= Angular Displacement = Angular Distance h = Cloud-Top Height
H = Satellite Height
O
S
A
hC N
B
R
H
S
O
B
C
C'll
N
O'
B'B Lx
Lxf
Ly
fB
dLy = R (rd) (km)
li(rd) = [hc (H+R) Sin(dli)] /{R[(H+R) Cos(dli) – (R+ hc)]}
dLx = R λ (rd) (km)
h = 20 kmh = 19 kmh = 18 kmh = 17 kmh = 16 kmh = 15 kmh = 14 kmh = 13 kmh = 12 kmh = 11 kmh = 10 kmh = 9 kmh = 8 kmh = 7 kmh = 6 kmh = 5 kmh = 4 kmh = 3 kmh = 2 kmh = 1 km
0 5 10 15 20 25 30 35 40 45 50 55 60 65
50
45
40
35
30
25
20
15
10
5
0
Spa t
ial D
isp l
acem
e nt (
km)
Longitudinal or Latitudinal Distance from a Geostationary Satellite (Degrees)
Cloud Top Relief SD for GOES Images
Clo
u d-T
op H
eig h
t (km
)
Relationship between Cloud Distance from Satellite, Cloud-Top Height and related Spatial Displacement
Stereoscopic Parallax Related to CTH
dp is the X- or Stereoscopic
Parallax Associated with hc,
Cloud-Top Height, between GOES-E and -W Images of a Cloud Element.
R = 6370 km (Radius of
Earth) Hs = 36000 km
IF: hc = 10 km
dp = 16.0 km
GOES-W
hc
GOES-E
60° X
Hs
(75W
)
(135W)
Rdp
Earth
C
250
250
215
215
215
215 K250 K
265
225
225
265
225K265K
Latit
ude
(Pix
el n
umbe
rs)
Longitude (Pixel Numbers)Comparison between x-parallaxes related to 2, cold (blue) and warm (red), temperaturesbetween IR cloud-top BT of corresponding pairs of GOES-E (solid line) and GOES-W (dotted line) IR Images, over Southeastern U.S. (left) and Colorado (right).
Dependency of SD & CTH / BT La
titud
e (P
ixel
num
bers
)
Longitude (Pixel Numbers)
GOES-E (13) BT GOES-W (11) BT
X
Earth
GOES-W(135W)
GOES-E(75W)
BTE BTW
10 20 30 40 50
10
20
30
GOES-E Cloud-TopBrightness Temperature
10 20 30 40 50
10
20
30
hchc
X
Y
RR
Ground-Based Radar Rainfall
10
20
30
X Y 10 20 30 40 50
Cloud-Top Relief Spatial Displacement
(Pixel Size = 2 km)
GOES-W Cloud-TopBrightness Temperature
Designing a 3-pieces piecewise line with 6 parameters for Tb-Height/SD relationship. Optimized parameters are estimated based on minimizing X-parallax between a pair of simultaneous GOES-E and -W IR images using 3-D principle and Shuffled Complex Evolution algorithm (SCE-UA).
3-Piece Line (6 Parameters) 6 parameters of a 3-pieces Linear Profile
Designing Fitted Piecewise Cloud-top BT-Height Relationship Line
ParametersOptimized
Values
h0 2.6 km
T1 242 K
T2 225 K
l1 0.15 km/K
l2 0.105 km/K
l3 0.14 km/K
Longitude (Degrees, W
est)Latitude (Degrees, North)
CTH
(km
)Cloud-Top Height (CTH) 2-D and 3-D Maps
Estimated Cloud-Top Height
3-D cloud-top height map
Estimated cloud-top height and their related contours (right) in the forms of 2-D and 3-D maps.
Clo
ud-to
p IR
Brig
htne
ss T
empe
ratu
re (K
)
Original GOES-E (13) IR-BT Original GOES-W (11) IR-BT
Spatial Adjusted GOES-E IR-BT Spatial Adjusted GOES-W IR-BT
Cloud-top GOES-E & -W IR-BT Before and After Spatial Adjustment
Before Spatial Adjustment After Spatial Adjustment
Longitude (Degrees, West)
Latit
ude
(De g
rees
, Nor
t h)
- 45 - 30 - 15 0 15 30 45Difference Cloud-Top Brightness Temperature (Kelvin)
Longitude (Degrees, West)
Latit
ude
(De g
rees
, Nor
t h)
Comparison between differences of GOES-E (13) and GOES-W (11) cloud-top BT before (left) and after (right) adjusting cloud-top relief spatial displacement.
Difference of GOES-E and GOES-W IR-BT Before & After Spatial Adjustment
Before Adjustment
After Adjustment
minimum -32 -15
maximum 23 17
GO
ES
-W IR
Brig
htne
ss T
empe
ratu
re (K
)
GOES-E IR Brightness Temperature (K)
Corr. = 0.95RMSE = 4.65
Pixel based GOES-11 (-W) IR cloud-top BT versus GOES-13 (-E) IR before (left) and after (right) adjustment of cloud-top relief spatial displacement.
GOES-W IR-BT vs. GOES-E Before and After Spatial Adjustment
Before Spatial Adjustment
GO
ES
-W IR
Brig
htne
ss T
empe
ratu
re (K
)
GOES-E IR Brightness Temperature (K)
Corr. = 0.987RMSE = 2.45
After Spatial Adjustment
Clo
ud-to
p IR
Brig
htne
ss T
empe
ratu
re (K
)
Original GOES-E (13) IR-BT Original GOES-W (11) IR-BT
Spatial Adjusted GOES-E IR-BT Spatial Adjusted GOES-W IR-BT
Cloud-top GOES-E & -W IR-BT Before and After Spatial Adjustment for Validation
GO
ES
-W IR
Brig
htne
ss T
empe
ratu
re (K
)
GOES-E IR Brightness Temperature (K)
Corr. = 0.96RMSE = 5.7
Pixel based GOES-11 (-W) IR cloud-top BT versus GOES-13 (-E) IR before (left) and after (right) adjustment of cloud-top relief spatial displacement using the derived CTH-IRBT for an independent study case, for validation.
GOES-W IR-BT vs. GOES-E Before and After Spatial Adjustment, Validation
Before Spatial Adjustment
GO
ES
-W IR
Brig
htne
ss T
empe
ratu
re (K
)
GOES-E IR Brightness Temperature (K)
Corr. = 0.98RMSE = 4.1
After Spatial Adjustment
Future Tasks• Evaluate the CTH estimates against CALIPSO-CTH (year 2);
• Investigate if the piecewise line model can be enhanced by increasing the number of parameters (pieces of lines) (Year 2);
• Test the model using other GOES thermal IR bands to investigate which IR frequency is more sensitive to cloud-top height, which will be used to improve CTH estimates (Year 2);
• Investigate and enhance Comparing the CTH estimates with the CTH from GOES-R ATBD algorithm and the CALIPSO-CTH product (Year 2);
• Investigate the impacts of seasonal, geographical, topography, storm types, and climate variability on model parameters and modify model for each (Year 3);
• Incorporate the 3-D based model into GOES-R ATBD model and evaluate the new CTH product against CALIPSO-CTH (Year 3);