1. Atmospheric Soundings, Surface Properties, Clouds The Bologna Lectures Paul Menzel NOAA/NESDIS/ORA

2. Relevant Material in Applications of Meteorological Satellites CHAPTER 6 - DETECTING CLOUDS 6.1 RTE in Cloudy Conditions6-1 6.2Inferring Clear Sky Radiances in Cloudy Conditions6-2 6.3Finding Clouds6-3 CHAPTER 7 - SURFACE TEMPERATURE 7.2. Water Vapor Correction for SST Determinations7-3 7.3Accounting for Surface Emissivity in the Determination of SST7-6 7.4Estimating Fire Size and Temperature7-6 CHAPTER 8 - TECHNIQUES FOR DETERMINING ATMOSPHERIC PARAMETERS 8.1 Total Water Vapor Estimation8-1 8.3 Cloud Height and Effective Emissivity Determination8-8 8.6 Satellite Measurement of Atmospheric Stability8-13

3. Earth emitted spectra overlaid on Planck function envelopes CO2 H20 O3 CO2

5. Profile Retrieval from Sounder Radiances p s I =  sfc B (T(p s ))  (p s ) -  B (T(p)) [ d  (p) / dp ] dp. o I1, I2, I3,...., In are measured with the sounder P(sfc) and T(sfc) come from ground based conventional observations  (p) are calculated with physics models First guess solution is inferred from (1) in situ radiosonde reports, (2) model prediction, or (3) blending of (1) and (2) Profile retrieval from perturbing guess to match measured sounder radiances

6. Sounder Retrieval Products Direct brightness temperatures Derived in Clear Sky 20 retrieved temperatures (at mandatory levels) 20 geo-potential heights (at mandatory levels) 11 dewpoint temperatures (at 300 hPa and below) 3 thermal gradient winds (at 700, 500, 400 hPa) 1 total precipitable water vapor 1 surface skin temperature 2 stability index (lifted index, CAPE) Derived in Cloudy conditions 3 cloud parameters (amount, cloud top pressure, and cloud top temperature) Mandatory Levels (in hPa) sfc78030070 100070025050 95067020030 92050015020 85040010010

7. Remote Sensing Regions Windows to the atmosphere (regions of minimal atmospheric absorption) exist near 4  m and 10  m; these are used for sensing the temperature of the earth surface and clouds. CO 2 absorption bands at 4.3  m and 15  m are used for temperature profile retrieval; because these gases are uniformly mixed in the atmosphere in known portions they lend themselves to this application. The water vapor absorption band near 6.3  m is sensitive to the water vapor concentration in the atmosphere; as H20 is not uniformly mixed in the atmosphere, measurements in this spectral region are used to infer moisture distribution in the atmosphere. The ozone absorption band at 9.7  m reveals locations of O3 concentration in the upper atmosphere.

9. GOES Sounder Spectral Bands: 14.7 to 3.7 um and vis

10. GOES Imager -- Spectral Coverage

11. Comparison of GOES-8 PW with microwave retrievals GOES Sounder total precipitable water (PW) values compare well with co- located microwave radiometer measurements at the CART site (Lamont, OK). “Flat” first-guess trace is adjusted by sounder radiances to capture the trend and range of total moisture.

12. Comparison of GOES-8 PW with microwave retrievals A scatter plot comparing MWR integrated water vapor values and GOES- 8 first guess/physical retrieval values at the CART site. RMS and bias values for all matches are quantified in the lower right hand corner.

14. Interactive Viewing of GOES Sounder DPI Time Series UW/Madison/CIMSS NOAA/NESDIS/ORA/ARAD/ASPT A new interactive web site (http://cimss.ssec.wisc.edu/goes/realtime/gdpiviewer.html) allows users to view time series of GOES derived product imagery (Lifted Index (LI), Precipitable Water (PW), and Convective Available Potential Entergy (CAPE)) by clicking on a desired location within the latest Derived Product Image (DPI). (graph lines: thick when sounding available, thin when cloud obscured) (example from Dec. 2, 1999) (de-stabilizing with time)

15. Detecting Clouds (IR) IR Window Brightness Temperature Threshold and Difference Tests IR tests sensitive to sfc emissivity and atm PW, dust, and aerosols BT11 < 270 BT11 + aPW * (BT11 - BT12) < SST BT11 + bPW * (BT11 - BT8.6) < SST aPW and bPW determined from lookup table as a function of PW BT3.9 - BT11 > 12 indicates daytime low cloud cover BT11 - BT12 < 2 (rel for scene temp) indicates high cloud BT11 - BT6.7 large neg diff for clr sky over Antarctic Plateau winter CO2 Channel Test for High Clouds BT13.9 < threshold (problems at high scan angle or high terrain)

17. Detecting Clouds (vis) Reflectance Threshold Test r.87 > 5.5% over ocean indicates cloud r.66 > 18% over vegetated land indicates cloud Near IR Thin Cirrus Test r1.38 > threshold indicates presence of thin cirrus cloud ambiguity of high thin versus low thick cloud (resolved with BT13.9) problems in high terrain Reflectance Ratio Test r.87/r.66 between 0.9 and 1.1 for cloudy regions must be ecosystem specific Snow Test NDSI = [r.55-r1.6]/ [r.55+r1.6] > 0.4 and r.87 > 0.1 then snow

19. aa MODIS cloud mask example (confident clear is green, probably clear is blue, uncertain is red, cloud is white) 1.6 µm image0.86 µm image11 µm image3.9 µm imagecloud mask Snow test (impacts choice of tests/thresholds) VIS test (over non-snow covered areas) 3.9 - 11 BT test for low clouds 11 - 12 BT test (primarily for high cloud) 13.9 µm high cloud test (sensitive in cold regions)

20. AVIRIS Movie #2 AVIRIS Image - Porto Nacional, Brazil 20-Aug-1995 224 Spectral Bands: 0.4 - 2.5  m Pixel: 20m x 20m Scene: 10km x 10km

21. MODIS identifies cloud classes

22. Clouds separate into classes when multispectral radiance information is viewed

23. Multispectral data reveals improved information about ice / water clouds

24. Cloud Composition Ice Cloud Water Cloud Contrails Infrared Temperature Difference - 8.6  m (Band 29) - 11.0  m (Band 31) Infrared Temperature Difference - 11.0  m (Band 31) - 12.0  m (Band 32) Image Over Kansas - 21 April 1996

25. Tri-spectral IR thermodynamic phase algorithm 8.6-11 vs 11-12 when slope > 1 then ice when slope < 1 then water ice cloud April 1996 Success water cloud Jan 1993 TOGA/ COARE Strabala, Menzel, and Ackerman, 1994, JAM, 2, 212-229. Baum et al, 2000, JGR, 105, 11781-11792.

26. Water phase clouds with 238K < T c < 253K

29. RTE in Cloudy Conditions I λ = η I cd + (1 - η) I c where cd = cloud, c = clear, η = cloud fraction λ λ o I c = B λ (T s )  λ (p s ) +  B λ (T(p)) d  λ. λ p s p c I cd = (1-ε λ ) B λ (T s )  λ (p s ) + (1-ε λ )  B λ (T(p)) d  λ λ p s o + ε λ B λ (T(p c ))  λ (p c ) +  B λ (T(p)) d  λ p c ε λ is emittance of cloud. First two terms are from below cloud, third term is cloud contribution, and fourth term is from above cloud. After rearranging p c dB λ I λ - I λ c = ηε λ   (p) dp. p s dp Techniques for dealing with clouds fall into three categories: (a) searching for cloudless fields of view, (b) specifying cloud top pressure and sounding down to cloud level as in the cloudless case, and (c) employing adjacent fields of view to determine clear sky signal from partly cloudy observations.

30. Cloud Clearing For a single layer of clouds, radiances in one spectral band vary linearly with those of another as cloud amount varies from one field of view (fov) to another Clear radiances can be inferred by extrapolating to cloud free conditions. R CO2 R IRW cloudy clear x xx x x partly cloudy N=1N=0

31. Paired field of view proceeds as follows. For a given wavelength λ, radiances from two spatially independent, but geographically close, fields of view are written I λ,1 = η 1 I λ,1 cd + (1 - η 1 ) I λ,1 c, I λ,2 = η 2 I λ,2 cd + (1 - η 2 ) I λ,2 c, If clouds are at uniform altitude, and clear air radiance is in each FOV I λ cd = I λ,1 cd = I λ,2 cd I λ c = I λ,1 c = I λ,2 c cd c c η 1 (I λ - I λ ) η 1 I λ,1 - I λ = = η* =, cd c c η 2 (I λ - I λ ) η 2 I λ,2 - I λ where η* is the ratio of the cloud amounts for the two geographically independent fields of view of the sounding radiometer. Therefore, the clear air radiance from an area possessing broken clouds at a uniform altitude is given by c I λ = [ I λ,1 - η* I λ,2 ] /[1 - η*] where η* still needs to be determined. Given an independent measurement of surface temperature, T s, and measurements I w,1 and I w,2 in a spectral window channel, then η* can be determined by η* = [I w,1 - B w (T s )] / [I w,2 - B w (T s )] and I λ c for different spectral channels can be solved.

32. Cloud Properties RTE for cloudy conditions indicates dependence of cloud forcing (observed minus clear sky radiance) on cloud amount (  ) and cloud top pressure (p c ) p c (I - I clr ) =    dB. p s Higher colder cloud or greater cloud amount produces greater cloud forcing; dense low cloud can be confused for high thin cloud. Two unknowns require two equations. p c can be inferred from radiance measurements in two spectral bands where cloud emissivity is the same.  is derived from the infrared window, once p c is known. This is the essence of the CO2 slicing technique.

33. Moisture Moisture attenuation in atmospheric windows varies linearly with optical depth. - k u  = e = 1 - k u For same atmosphere, deviation of brightness temperature from surface temperature is a linear function of absorbing power. Thus moisture corrected SST can inferred by using split window measurements and extrapolating to zero k T s = T bw1 + [ k w1 / (k w2 - k w1 ) ] [T bw1 - T bw2 ]. Moisture content of atmosphere inferred from slope of linear relation.

34. Early SST algorithms * IRW histogram of occurrence f of observed brightness temperatures T f(T) = f s exp [ -(T - T sfc ) 2 /2  2 ] instrument noise / scene variability  produce Gaussian distribution; warm side of histogram reveals T s = T (d2f/dT2=0) -  * Three point method combinations of (T i, f i ), (T j, f j ), and (T k, f k ) on the warm side of the histogram enable 3 equations / 3 unknowns, hence a histogram of T s solutions. * Least squares method ln (f(T)) = ln (f s ) - T s 2 /2  2 + T s T/  2 - T 2 /2  2 has the form ln (f(T)) = A o + A 1 T + A 2 T 2, so T s = - A 1 /(2A 2 ).

35. Histograms of infrared window brightness temperature in cloud free and cloud contaminated conditions

36. Water Vapor Correction for SST Determinations * Water vapor correction (  T) T s = T b +  T ranges from 0.1 C in cold/dry to 10 K in warm/moist atmospheres for 11 um IRW observations. * Water vapor correction is highly dependent on wavelength. * Water vapor correction depends on viewing angle. * In IRW for small water vapor concentrations,  w = e -K w u ~ 1 - K w u so that T s = T bw1 + [ K w1 / (K w2 - K w1 ) ] [T bw1 - T bw2 ]. linear extrapolation to moisture free atmosphere * Regression of clear sky IRW obs and collocated buoys create current operational algorithm T s =A0+A1* T bw1 +A2*( T bw1 -T bw2 )+A3*(T bw1 -T bw2 ) 2 a quadratic term helps account for occasional large water vapor concentrations.

37. Cloud Detection Several multispectral methods have evolved to detect clouds in the area of interest. Input data are vis, T 3.9, T 11, and T 12, T 11 @  30min, and SST guess. General tests include: T 11 > 270 Kocean rarely frozen T 11 > T 12 + 4 Kclouds affect moisture correction vis < 4%clouds reflect more than ocean sfc T 11 - T 3.9 > 1.5 Ksubpixel clouds  T 11 < 0.3 K  SST over 1 hr small -2 K < SST- guess < 5 K  SST over days bounded

38. Advantages of Geostationary SST Estimates *10 times more observations of a given location *multispectral cloud detection supplemented by temporal persistence checks *clear sky viewing enhanced by persistence (e.g. can wait for clouds to move through) *daily composite provides good spatial coverage *can discern diurnal excursions in SST *can track SST motions as estimates of ocean currents

39. GOES daily composite SST reveals small scale features in oceans

40. GOES detects diurnal SST excursions of 2-3 C in calm waters

41. Accounting for surface emissivity When the sea surface emissivity is less than one, there are two effects that must be considered: (a) the atmospheric radiation reflects from the surface, and (b) the surface emission is reduced from that of a blackbody. The radiative transfer can be written p s I =  B (T s )  (p s ) +  B (T(p))d  (p) o p s + (1-  )  (p s )  B (T(p))d  ‘ (p) o where  ‘ (p) represents the transmittance down from the atmosphere to the surface. This can be rewritten I =  B (p s )  (p s ) + B (T A )[1 -   (p s ) -  (p s ) 2 +   (p s ) 2 ]. Note that as the atmospheric transmittance approaches unity, the atmospheric contribution expressed by the second term becomes zero.

42. HIS and GOES radiance observations plotted in accordance with the radiative transfer equation including corrections for atmospheric moisture, non-unit emissivity of the sea surface, and reflection of the atmospheric radiance from the sea surface. Radiances are referenced to 880 cm-1. The intercept of the linear relationship for each data set represents a retrieved surface skin blackbody radiance from which the SST can be retrieved.

43. Comparison of ocean brightness temperatures measured by a ship borne interferometer (AERI), by an interferometer (HIS) on an aircraft at 20 km altitude, and the geostationary sounder (GOES-8). Corrections for atmospheric absorption of moisture, non-unit emissivity of the sea surface, and reflection of the atmospheric radiance from the sea surface have not been made.

45. GOES 3 by 3 FOVs (30 km) MODIS 5 by 5 FOVs (5 km) 11 micron

46. 30km resolution GOES 500hPa T 5km resolution MODIS

47. MODIS 2000/09/05-08 Daytime 850 hPa Temperature (K) for 4 days

48. NOAA-15 AMSU-A 2000/09/05 Daytime 850 hPa Temperature (K) for one day

49. MODIS

52. GOES vs. MODIS 2000/06/30 1600 UTC Total Precipitable Water (mm)

53. GOES 30 km resolution MODIS 5 km resolution TPW GOES vs. MODIS 2000/06/30 1600 UTC Total Precipitable Water (mm)

54. MODIS total precipitable water vapor shows a wet bias wrt GOES; bias 1.5 mm and rms of 3 mm; bias will be removed after more validation

55. MODIS 2000/09/05-08 Daytime Total Precipitable Water (cm) values over land not shown to facilitate comparison with AMSU

56. NOAA-15 AMSU-A 2000/09/05 Daytime Total Precipitable Water (mm)

57. MODIS TPW: Upper panel; AMSU TPW: Lower panel

58. GOES 30 km resolution MODIS 5 km resolution Ozone

59. MODIS ozone is very close to the GOES ozone (over North America); rms of about 10 Dobsons; polar extreme ozone values will be improved

60. MODIS 2000/09/05-08 Daytime Total Ozone (Dobsons)

61. Earth Probe TOMS 2000/09/05 Total Ozone (Dobsons)

62. MODIS Ozone TOMS Ozone

63. Early Estimates of UW MODIS Product Quality MODIS IR window radiances agree to within 1 C with GOES and ER-2 MAS/SHIS Cloud mask has demonstrated advantages of new multispectral approach sun glint, desert, and polar problems diminished MODIS cloud top pressures compare well with HIRS; aircraft validation is better than 50 mb. MODIS cloud phase determinations are revealing interesting patterns; first ever global day/night ice/water cloud determinations; validations pending. MODIS tropospheric temperatures compare well with AMSU; rms better than 1 C, both within 2 C of radiosonde observations MODIS total precipitable water vapor shows a wet bias wrt GOES; bias 1.5 mm and rms of 3 mm; bias will be removed after more validation. MODIS ozone is very close to the GOES ozone (over North America); rms of about 10 Dobsons; polar extreme ozone values will be improved