SHORT RANGE STRUCTURE OF CLOUDS Stephen E. Schwartz and Dong Huang
Upton, Long Island, NY
Scales and Scaling in the Climate System:
Bridging theory, climate models and data Jouvence Centre, Quebec
October 4-7, 2015
www.ecd.bnl.gov/steve
WHY ARE WE INTERESTED IN CLOUDS? • Clouds exert strong radiative influence in shortwave (– 48 W
m-2) and longwave (+ 17 W m-2) global average; much more locally and instantaneously. Even thin clouds exert strong radiative effects.
• Need to accurately represent clouds in weather and climate models.
• Any change in clouds could augment or diminish the climate impact of increasing greenhouse gases – cloud feedbacks.
• Clouds may be visible manifestation of atmospheric dynamics and variability.
• Cloud-aerosol interactions and radiative forcing of climate change.
WHAT IS A CLOUD?AMS Glossary of Meteorology (2000)
A visible aggregate of minute water droplets and/or ice particles in the atmosphere above the earth’s surface.Total cloud cover: Fraction of the sky hidden by all visible clouds.
Clothiaux, Barker, & Korolev (2005)
Surprisingly, and in spite of the fact that we deal with clouds on a daily basis, to date there is no universal definition of a cloud. . . . Ultimately, the definition of a cloud depends on the threshold
sensitivity of the instruments used.
Ramanathan, JGR (ERBE, 1988)
Cloud cover is a loosely defined term.
Potter Stewart (U.S. Supreme Court, 1964)
I shall not today attempt further to define it, but I know it when I
see it.
CLOUD OPTICAL DEPTH, Vertical integral of scattering coefficient .Scattering coefficient is probability of photon being
scattered per unit distance.For cloud drops (radius r >> wavelength of light),
= 2 r2 n (n is number concentration of cloud drops).Optical depth is commonly used measure of radiative
influence of a cloud.Thick clouds, > 100: Almost all light is scattered
upward. Thin clouds, < 1 : Most light is transmitted.
~
~
CLOUD RADIATIVE EFFECTDependence on shortwave optical depth and cloud-top temperature
24-Hour average CRE, north central Oklahoma, at equinox
-200
-100
0
100
0.01 0.1 1 10 100
Cloud shortwave optical depth
Shortwave
Tct 280 260 240 220 K
Longwave
Net
-200
-100
0
100
200
0 02 04 06 08 001-10
-5
0
5
10
0 20.0 40.0 60.0 80.0 01.0
-60
-40
-20
0
20
40
60
0.0 2.0 4.0 6.0 8.0 0.1
-200
-100
0
100
200
0 2 4 6 8 01
Cloud shortwave optical depth
Net CRE depends on optical depth and cloud-top temperature even in sign.
Clo
ud
Ra
dia
tive
Effe
ct
at
To
p o
f A
tmo
sp
he
re , W
m-2
MEASUREMENTS OF GLOBAL CLOUD FRACTION 0.8
0.7
0.6
0.5
Glo
bal m
ean
clou
d fra
ctio
n
Ocean
Land
Global
Warren et al. Atlases
CALIPSO-STTOVS-PathBHIRS-NOAA
MODIS-STAIRS-LMDISCCPPATMOS-XMISRMODIS-CEATSR-GRAPEPOLDER
CALIPSO-GOCCP
Time Modified from Stubenrauch, Rossow, ... Ackerman, ... Chepfer, DiGirolamo, ... Winker et al., BAMS, 2013
• For clouds with optical depth > 0.1 global cloud fraction is about 68%. • Cloud fraction increases to 73% when including subvisible cirrus with
optical depth down to 0.01 (e.g. CALIPSO) and decreases to about 56% for clouds with optical depth > 2 (e.g. POLDER).
• Key reasons for differences: resolution and threshold.
CHALLENGES IN CHARACTERIZING CLOUDS AND REPRESENTING THEM IN MODELS
• Ephemeral. Clouds are tenuous, hard to define, harder to study.
Condensed cloudwater is about 1% of surrounding water vapor.
Cloudwater amount is highly dependent on condensation or evaporation associated with cloud vertical motion.
• Multiple scales. Clouds exhibit structures on many scales, from thousands of kilometers down to 1 meter.
New methods of characterizing clouds are welcome.
CLOUD PHOTOGRAPHY FROM SPACE
LOOKING DOWNWARD
EARTH FROM 1.5 MILLION KILOMETERS DSCOVR satellite at Earth–Sun Lagrange point
July 6, 2015; Credit – NASA
2048 ! 2048 pixels; nadir resolution 8 km.
EARTH FROM 1.5 MILLION KILOMETERS DSCOVR satellite at Earth–Sun Lagrange point
July 6, 2015; Credit – NASA
Focus in on a box over the southeastern Pacific.
SOUTHEASTERN PACIFIC AT NATIVE RESOLUTION
Rich structure in clouds at a variety of scales.
RED IMAGE
Red image emphasizes clouds (red filter in black and white photography).
RED/(RED + BLUE) IMAGE – RRB
Red/(Red + Blue) normalizes for intensity and brings out thin clouds.
FALSE COLOR RED/(RED + BLUE) IMAGE
1.00.80.60.40.20.0
Clou
d fra
ction
0.500.450.400.350.30
Red/(Red+ Blue)
8
6
4
2
0
Red/(Red + Blue) normalizes red radiance and brings out thin clouds. Pixels with RRB greater than ~ 0.45 are pretty confidently cloud; cloud
fraction, evaluated from integral of PDF, ~32%. Values of RRB less than ~ 0.35 are pretty confidently cloud free; cloud
fraction ~ 73%. Cloud fraction ranges from 32% to 73%. There is no value of RRB that uniquely separates cloud from clear sky.
ZOOM IN ON POCKETS OF CLOSED CELLS
Organized convective cells contributing to planetary reflectance.
FALSE COLOR RED/(RED + BLUE) IMAGE
Note organized structure in RRB image.
CLOUD PHOTOGRAPHY FROM THE SURFACE LOOKING UPWARD
3 Color, RGB, 16bit1200 mm focal length
(35 mm equiv)1 Pixel = 6 rad (20 rad)FOV 22 29 mrad
(2 3 sun diameters)ProgrammableWi-Fi communicationWeather resistant$400
HIGH RESOLUTION IMAGER Fujifilm FinePix S114 Megapixels 3456 4608
1200 mm EQUIVALENT FOCAL LENGTH
Todd Vorenkamp, B&H Photo, NYC
That's 1.2 meters!
NARROW FIELD OF VIEW
29 ! 22 mrad ! 3 ! 2 sun (or moon) diameters, 29 ! 22 m at 1 km
RESOLUTION
OBSERVATION GEOMETRY
0.5
1.0
1.5
2 radians
East West
South
North
06:00
08:00
10:00
12:00 14:0016:00
18:00
Sun, angular diameter
Both drawn 10 times
actual angular dimension
Solar zenith angle and
azimuth, July 15, 2014,
6 am to 6 pm EDT
Zenith looking
Upton, Long Island, NY
STRENGTHS AND ADVANTAGES Black background of outer space: No surface effects (to first order). No side-wall issues; no correction sky cover to ground cover. Readily available data acquisition hardware and software. Available, easy-to-use image analysis and processing software. Lots of data!
WEAKNESSES AND LIMITATIONS Two-dimensional only. Daytime only. Limited wavelength range. Small fraction of sky; extremely local. Aerosol masquerades as thin cloud. Lots of data!
ZENITH RADIANCE AND RED/(RED + BLUE) Dependence on cloud optical depth and solar zenith angle
Cos SZA Red Blue 1.0 0.8 0.5 0.2
0.8
0.6
0.4
0.2
0.0
Norm
alize
d ze
nith
radia
nce
Cos SZA Red Blue 1.0 0.8 0.5 0.2
1501251007550250Cloud optical depth
Cos SZA 1.0 0.8 0.5 0.2
0.55
0.50
0.45
0.40
0.35
0.30
0.25
0.20
Red/
(Red
+ B
lue)
543210Cloud optical depth
Cos SZA 1.0 0.8 0.5 0.2
Zenith radiance increases rapidly with !, peaks, and then decreases. RRB increases rapidly and then plateaus independent of solar zenith angle.
0.55
0.50
0.45
0.40
0.35
0.30
0.25
Red/(
Red +
Blu
e)
17:15 17:30 17:45 18:00
353 374 395
416437
458 479 500 521
542
563
571
333
40
30
20
10
0
Inte
gra
ted lid
ar
retu
rn 2
-3 k
m Photo #
Lidar over same 1 hr period; superimposed integrated lidar signal
and mean Red/(Red + Blue), RRB, from photos
442
447
Thin, single cloud layer, base ~ 2.6 km.
Note high RRB for cloud; low for clear.
Note high RRB even for very thin cloud, e.g. 442-447.
Could be broken cloud (442) or thin uniform cloud (443-447) more likely.
Note strong effect on RRB despite low optical depth inferred from transmittance.
CCNY, NYC, 2015-05-22
SPATIAL AND TEMPORAL VARIABILITY239 Successive photos at 15 s intervals over 1 hour, 2015-05-22, CCNY, NYC
333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353
354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374
375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395
396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416
417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437
438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458
459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479
480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500
501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521
522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542
543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563
564 565 566 567 568 569 570 571
Note high variability at 15 s intervals.
Note high spatial variability; one image ~ 57 X 75 m
SPATIAL AUTOCORRELATION Broken single cloud, 2.6 km; NYC CCNY 2015-05-22
• No meaningful cloud fraction. • Spatial autocorrelation reasonably symmetric with length ~ 15 m.
SPATIAL AUTOCORRELATION Dependence on location within single broken cloud
• Autocorrelation structure is highly variable within image.
GOES VISIBLE CHANNEL, SGP, 2015-0731, 1600 UTC
Thin broken clouds visible in 1-km resolution GOES image.
SIX MINUTES IN OKLAHOMA, JULY 31, 2015
16:12:48 16:12:52 16:12:56 16:13:00 16:13:04 16:13:08 16:13:12 16:13:16 16:13:20 16:13:24 16:13:28 16:13:32 16:13:36 16:13:40 16:13:44
16:13:48 16:13:52 16:13:56 16:14:00 16:14:04 16:14:08 16:16:12 16:16:16 16:14:20 16:14:24 16:14:28 16:14:32 16:14:36 16:14:40 16:14:44
16:14:48 16:14:52 16:14:56 16:15:00 16:15:04 16:15:08 16:15:12 16:15:16 16:15:20 16:15:24 16:15:28 16:15:32 16:15:36 16:15:40 16:15:44
16:15:48 16:15:52 16:15:56 16:16:00 16:16:04 16:16:08 16:16:12 16:16:16 16:16:20 16:16:24 16:16:28 16:16:32 16:16:36 16:16:40 16:16:44
16:16:48 16:16:52 16:16:56 16:17:00 16:17:04 16:17:08 16:17:12 16:17:16 16:17:20 16:17:24 16:17:28 16:17:32 16:17:36 16:17:40 16:17:44
16:17:48 16:17:52 16:17:56 16:18:00 16:18:04 16:18:08 16:18:12 16:18:16 16:18:20 16:18:24 16:18:28 16:18:32 16:18:36 16:18:40 16:18:44
N
S
E W
167 m @ 1410 m
Note general eastward motion with time; note also evolution of features.
SIX MINUTES IN OKLAHOMA, JULY 31, 2015 N
S
E W
30 m @ 1410 m
16:12:48 16:12:52 16:12:56 16:13:00 16:13:04 16:13:08 16:13:12 16:13:16 16:13:20 16:13:24 16:13:28 16:13:32 16:13:36 16:13:40 16:13:44
16:13:48 16:13:52 16:13:56 16:14:00 16:14:04 16:14:08 16:16:12 16:16:16 16:14:20 16:14:24 16:14:28 16:14:32 16:14:36 16:14:40 16:14:44
16:14:48 16:14:52 16:14:56 16:15:00 16:15:04 16:15:08 16:15:12 16:15:16 16:15:20 16:15:24 16:15:28 16:15:32 16:15:36 16:15:40 16:15:44
16:15:48 16:15:52 16:15:56 16:16:00 16:16:04 16:16:08 16:16:12 16:16:16 16:16:20 16:16:24 16:16:28 16:16:32 16:16:36 16:16:40 16:16:44
16:16:48 16:16:52 16:16:56 16:17:00 16:17:04 16:17:08 16:17:12 16:17:16 16:17:20 16:17:24 16:17:28 16:17:32 16:17:36 16:17:40 16:17:44
16:17:48 16:17:52 16:17:56 16:18:00 16:18:04 16:18:08 16:18:12 16:18:16 16:18:20 16:18:24 16:18:28 16:18:32 16:18:36 16:18:40 16:18:44
Note temporal evolution and spatial inhomogeneity even at this scale.
MULTIPLE MEASURES OF CLOUD North central Oklahoma, 2015-07-31
Lidar gives cloud base height and measure of transmittance. Integrated lidar return is measure of cloud amount. R/(R+B) from two zenith pointing cameras: 21 × 29 mrad, 120 × 160 mrad. 673/(673 + 440) from zenith pointing spectral radiometer. All measures show more or less coherent signals of intermittent clouds.
COMPARE WFOV AND NFOV IMAGES Examine subset of WFOV corresponding to NFOV image
Red/(Red + Blue) false-color images and histogram are virtually identical
for two images, showing robustness of quantity.
STRENGTHS OF RED/(RED + BLUE) AS MEASURE OF CLOUDINESS
• Continuously variable quantity, as opposed to binary numerator in evaluating cloud fraction.
• Nearly monotonic in cloud optical depth !, with plateau at large !.
• Nearly independent of solar zenith angle, except at low !.
• Nearly independent of scene brightness.
• Suitable for spatial and temporal averaging as R / R+ B .
CONCLUSIONS • Clouds can be imaged with resolution better than 1 meter by high
resolution photography from the surface.
• Clouds frequently exhibit high spatial variability on scales of meters to tens of meters.
• Even very thin clouds are readily detected by color and quantified as Red/(Red + Blue).
• Red/(Red + Blue) varies greatly for thin clouds. In scenes with variable cloudiness cloud fraction cannot be uniquely defined.
• We are developing tools to model the radiance of such clouds and infer distributions of cloud optical depth and model cloud radiative effect.
• We welcome any suggestions for analysis of cloud spatial properties from photographic images.
