GLOBAL CIRRUS CLOUDS AND UPPER TROPOSPHERIC …



HIRS Observations of a Decline in High Clouds Since 1995

W. Paul Menzela and Donald P. Wylieb

a Office of Research and Applications , NOAA / NESDIS,

Madison,WI, USA

b Cooperative Institute for Meteorological Satellite Studies (CIMSS)

University of Wisconsin -Madison , Madison, WI, USA

Abstract

Seasonal changes in semi-transparent or cirrus global cloud cover have been monitored with multi-spectral observations with the polar orbiting HIRS (High resolution Infrared Radiation Sounder) since June 1989. Trends in cloud cover, inferred from monthly averages of the HIRS cloud observation frequencies, indicate that there has been a decrease of more than 5% in the tropics since January 1995. In the northern mid-latitudes a decrease of more than 8% is evident. These decreases are mainly observed over land; ocean cloud cover is mostly stable.

1. UW HIRS CIRRUS CLOUD TRENDS

The HIRS measurements in the carbon dioxide absorption band at 15 microns are used to detect cloud and calculate both cloud top pressure and effective emissivity from radiative transfer principles. The technique and details of its application with HIRS data are described in Wylie et al. (1994) and Wylie and Menzel (1999). The HIRS data have a higher sensitivity to semi-transparent cirrus clouds than visible and infrared window techniques; the threshold for detection appears to be at IR optical depths greater than 0.05 (in general visible optical depths are twice the IR optical depth). Seasonal changes in global high cloud cover are well characterized by these CO2 sensitive measurements.

Global UW HIRS cloud statistics since June 1989 are shown in Table 1. They are separated by cloud type into clear sky ((IR < 0.05), thin ((IR < 0.7), thick ((IR > 0.7), and opaque ((IR > 3.0) clouds and separated by level in the atmosphere above 6 km, between 3 and 6 km, and below 3 km. On the average for summer and winter, HIRS finds thin clouds in 23% of all observations, thick clouds in 22%, and opaque clouds in 29%. Thus these HIRS observations imply that clouds are found in 74% of all HIRS observations from 65 S to 65 N.

Trends in cloud cover are inferred from monthly averages of the cloud frequencies for each satellite and plotted as one continuous record (see Figure 1). The 12 year record of cloud frequency from NOAA HIRS data shows cloud cover decreases since 1995 in all zones; the northern hemisphere shows the most and the tropics show the least. These decreases are mainly observed over land; ocean cloud cover is mostly stable (Figure 2). There is a large diurnal signature in the detection of clouds over land that must be considered when viewing Figure 2.

2. INDIVIDUAL SATELLITE CLOUD DETECTION TRENDS

Detection of clouds by individual satellites is presented in Figure 3. The afternoon (2 pm local time) satellites, NOAA 11 and 14 show cloud increases over the years that are likely due to the change in satellite overpass; each satellite orbit drifted 3 to 4 hours later into the afternoon during the five plus years of operation (see Figure 4). More clouds are found later in the afternoon over land (Schreiner et al., 2001). NOAA 16, the current 2 pm satellite, shows low cloud cover and is maintaining a 2-3 pm local solar time overpass. However the sunrise-sunset satellites, NOAA 10 and 12, show a big drop early in 1995 that cannot be attributed to orbit drift; both maintain their local overpass times within one hour. NOAA 11 appearing as a sunrise-sunset satellite in 1998 supports this drop as does NOAA 15.

3. SENSITIVTY TO ANCILLARY SURFACE TEMPERATURE

The cloud mask, used as input to the CO2 Slicing estimates of cloud properties, relies on a comparison of water vapor corrected infrared window brightness temperatures with an a priori estimate of surface temperature. If these are within 2 degrees Centigrade of each other the field of view (FOV) is said to be cloud free. The radiances from the cloud free FOVs are then interpolated into cloudy areas to create a global map of cloud free radiances. FOVs declared cloudy in the cloud mask can subsequently be declared cloud free if the observed radiances and the interpolated cloud free estimates are within 1 mW/m2/ster/cm-1. If the surface temperature used in the cloud mask algorithm is too warm in a given region, most FOVs in this region will initially be considered cloudy. The clear radiance estimates for this region (acquired by interpolation of clear FOV radiance values from another region) will be colder; thus the observed minus clear radiance differences will be small and the cloud algorithm reclassifies the warm region FOVs into the clear category.

Sensitivity of the cloud detection algorithm to the a priori surface temperature was tested by varying the surface temperature by –2 to +2 degrees; Figure 5 shows cloud statistics change by 10% or more depending on latitude zone. If the surface temperature is within 1 degree, then the cloud statistics are more stable. When only high clouds are considered, there is even more stability.

Figure 6 shows the time trends of the HIRS clear radiance (brightness temperature) data. For the two satellites with time drift in their orbits (NOAA 11 and 14), the clear radiances decreased as the orbit drifted later into the afternoon. From our sensitivity test, we expect the cloud frequencies reported by these satellites to decrease. But the opposite occurred as NOAA 11 and 14 reported higher cloud frequencies. This supports the notion that the cloud change is not induced by a priori surface temperature difficulties but instead is likely a diurnal cloud increase in the afternoon.

4. TROPICAL CLOUD TRENDS

Recently, Wielicki et al (2002), using CERES (Clouds and the Earth’s Radiant Energy System) measurements of the Earth radiation budget, found that the observed broadband anomalies for LW-emitted, SW-reflected, and net radiation flux changes in the tropics are consistent with a decrease in tropical cloudiness. Figure 7 superimposes the Wielicki et al (2002) deviation of the solar reflected shortwave flux from the 1985-89 mean and the UW HIRS deviation from the twelve year mean of all and high cloud detection in the tropics (20 N to 20S). The decrease of tropical SW flux starts 1992 and the decrease of tropical cloud cover starts in 1995. While the start of the decrease differs, the trends in the two observation systems agree,

5. CONCLUSIONS

Twelve years trends in UW HIRS cloud statistics reveal that high clouds increased until 1995 and then gradually decreased below 1989 levels. Cloud cover decreases since 1995 occur in all zones (northern hemisphere, tropics, and southern hemisphere); the northern hemisphere shows the most and the tropics show the least. These decreases are mainly observed over land and especially in higher clouds; ocean cloud cover is mostly stable.

Differences in cloud statistics and their trends are consistent with changes in the local solar times of the orbits for the satellites that drifted. The satellite that did not drift, NOAA 12, shows a very large cloud frequency decrease.

A decrease in tropical cloud cover in the UW HIRS statistics is similar to that reported by Weilicki et al (2002) using CERES data.

6. Acknowledgments

This work was supported by NASA Contract NAG1-2196.

7. References

Ackerman, S. A., K. I. Strabala, W. P. Menzel, R. A. Frey, C. C. Moeller, and L. I. Gumley, 1998: Discriminating clear-sky from clouds with MODIS. J. Geophys. Res., 103, D24 , 32141-32158.

Schreiner, A. J., T. J. Schmit, and W. P. Menzel, 2001: Observations and trends of clouds based on GOES Sounder data. Jour. Geophys. Res.-Atmospheres. 106, 20349-20363.

Wielicki, B, A. T. Wong, R. P. Allan, A. Slingo, J. T. Kiehl, B. J. Soden, C. T. Gordon, A. J. Miller, S.-K. Yang, D. A. Randall, F. Robertson, J. Susskind, H. Jacobowitz, 2002: Evidence for Large Decadal Variability in the Tropical Mean Radiative Energy Budget. Science, 295, 841-844.

Wylie, D. P., and W. P. Menzel, 1999: Eight years of high cloud statistics using HIRS. J. Climate, 12, 170-184.

Wylie, D. P. W. P. Menzel, and K. I. Strabala, 1994: Four years of global cirrus cloud statistics using HIRS. J. Climate, 7, 1972-1986. __________________________________________________________________________________

Table 1: UW HIRS cloud statistics from 12 years of global near nadir HIRS measurements for the boreal summers and winters from June 1989 through August 2001 between 65 N and 65 S

| |Boreal Summer (June-August) |Boreal Winter (December-February) |

| |All Clouds |Thin Clouds |Thick Clouds |Opaque Clouds |All Clouds |Thin Clouds |Thick Clouds |Opaque Clouds |

|High >6km |36% |16% |15% |5% |36% |16% |16% |4% |

|Mid 3-6 km |25% | 9% | 11% |5% |30% | 9% |11% | 9% |

|Low 3.0. Comparisons with 1 km AVHRR data have indicated that the FOV is totally obscured by cloud when N( > 0.5 and 72% obscured by cloud when N( < 0.5. 6 km averaged 441 hPa in pressure height; 3 km averaged 665 hPa in pressure height.

__________________________________________________________________________________

[pic]

Figure 1. High cloud and all cloud observations found in HIRS measurements from June 1989 through August 2001. Polar orbiters NOAA 10 through 15 are part of this data set. NOAA 12 replaced NOAA 10 in September 1991, and NOAA 14 replaced NOAA 11 in April 1995. NOAA 15 replaced NOAA 12 in October 1998. Cloud observations over land from NOAA 11 are discarded because of the influence of the orbit drift.

[pic][pic]

Figure 2. The frequency of detection of all clouds (total cloud) and high cloud (above 440 hPa) over oceans (left) and over land (right) from the UW HIRS analyses.

[pic]

Figure 3. The frequency of detection of all clouds (total cloud) over land in the tropics (20 N to 20 S) by NOAA 10 through 16 since January 1989.

[pic]

Figure 4. Local overpass times for the NOAA satellites. Both NPOAA 11 and 14 exhibit orbit drift of three to four hours later into the afternoon during their five plus years of operation.

[pic][pic]

Figure 5. The sensitivity of all cloud (left) and high cloud (right) statistics to the ancillary surface temperature used in the moisture corrected infrared window cloud mask test.

[pic]

Figure 6. The clear radiance blackbody temperatures sensed by the HIRS on NOAA 10 through 16 in the tropics from 20 N to 20 S since January 1989.

[pic]

Figure 7. Comparison of the Wielicki et al (2002) deviation of the solar reflected shortwave flux from the 1985-89 mean and the UW HIRS deviation from the twelve year mean of all and high cloud detection in the tropics (20 N to 20S). The decrease of tropical SW flux starts 1992 and the decrease of tropical cloud cover starts in 1995.

................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download