SEARCH

SEARCH BY CITATION

Keywords:

  • precipitating clouds;
  • nonprecipitating clouds;
  • TRMM

Abstract

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Data and Methodology
  5. 3. Results
  6. 4. Conclusions
  7. Acknowledgments
  8. References
  9. Supporting Information

[1] Cloud properties retrieved from TRMM VIRS measurements are analyzed for PR-determined precipitating clouds (P-PCs) and nonprecipitating clouds (PN-PCs), respectively. The total cloud amount (CA) averaged across the tropics and subtropics in boreal summer is about 55.9 and 40.1, over ocean and land, respectively, with P-PCs contributing less than 10% to the total CA. Low P-PCs that have cloud top lower than 680 mb are extremely scanty, while low PN-PCs account for near half of total PN-PCs. The mean cloud optical thickness (COT) of P-PCs exceeds 60, approximately 10 times that of PN-PCs. According to ISCCP cloud classification, four primary cloud types of PN-PCs, cumulus, stratocumulus, altocumulus and cirrus are revealed, whereas deep convective clouds and cirrostratus are proved to be the first and second primary cloud type of P-PCs, implying a considerable amount of P-PCs with high cloud top but moderate COT.

1. Introduction

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Data and Methodology
  5. 3. Results
  6. 4. Conclusions
  7. Acknowledgments
  8. References
  9. Supporting Information

[2] Clouds have important impacts on the balance of energy in the Earth-Atmosphere system due to their interactions with solar and terrestrial radiation [Ramanathan, 1987; Ramanathan et al., 1989], which has been one of the most challenging aspects of climate change [Cess et al., 1989]. Designed to clarify the effects of various kinds of clouds on Earth's radiation budget and their potential impacts on climate change, the International Satellite Cloud Climatology Project (ISCCP) [Schiffer and Rossow, 1983] has compiled global cloud data sets based on measurements from a series of operational weather satellites for more than two decades. By using the ISCCP cloud products, studies have been motivated to characterize the properties and variability of various cloud types. Rossow and Schiffer [1999] presented a comprehensive report on global cloud properties by analyzing over 10 years' data sets, indicating a global mean total cloud amount (CA) 67.5 (expressed as a percentage), cloud top temperature (CTT) 261.5 K and cloud optical thickness (COT) 3.7.

[3] To stress the difference among various clouds, nine cloud types are designated in the ISCCP cloud classification scheme according to cloud top pressure (CTP) and cloud optical thickness (COT). Precipitating clouds (PCs), however, as one special cloud type, cannot be separated effectively from the cloud ensemble due to the inability of infrared and visible observations to identify instantaneous precipitation. The indirect relationship between CTT and precipitating particles results in that precipitation information derived from infrared/visible measurements is only acceptable in statistical sense, for which the GOES precipitation index (GPI) method can only work well in monthly rainfall estimation [Arkin and Xie, 1994]. Consequently, it is unable to get a reliable decision whether those pixels satisfying the ISCCP definition of PCs are actually precipitating or not, although these clouds are demonstrated to have both enough low CTT and enough ample cloud water path (CWP) [Lin and Rossow, 1997; Rossow and Schiffer, 1999]. Furthermore, the specific cloud properties of PCs, such as CA, CTT/CTP and COT, are unavailable so far.

[4] The nearly simultaneous measurements of Precipitation Radar (PR) and Visible and Infrared Scanner (VIRS) aboard the Tropical Rainfall Measurement Mission (TRMM) [Simpson et al., 1988] satellite provide a unique opportunity for accurately differentiating PCs from nonprecipitating clouds (N-PCs). Compared with previous cloud-target observations, the combined PR and VIRS measurements contain more sufficient information for a cloud view, with PR measurements supplying the instantaneous precipitation intensity, which were generally used as the criteria to detect PCs [Inoue and Aonashi, 2000; Masunaga et al., 2005; Fu et al., 2006; Fu and Liu, 2007]. For instance, Liu et al. [2007] examined the difference of the amount between PCs derived from infrared brightness temperature (TB11) and those detected by PR, denoting that only 35% (57%) of cold cloud area of TB11 less than 235 K (210 K) has PR-determined precipitation. In this paper, benefited from the merged PR and VIRS measurements, the universal difference of cloud properties between PCs and N-PCs was emphasized particularly. First, the global distribution of CA is investigated in terms of PCs and N-PCs, followed by a comparison of COT and CTP between these two categories of clouds. Then the PCs' occurrence frequency in the ISCCP cloud types is compared with that of N-PCs.

2. Data and Methodology

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Data and Methodology
  5. 3. Results
  6. 4. Conclusions
  7. Acknowledgments
  8. References
  9. Supporting Information

[5] TRMM standard products, 2A25 and 1B01, derived from TRMM PR and VIRS in three boreal summers from 1998 to 2000 are used in this study. Collocation of two data sets is made by merging adjacent VIRS pixels (5∼8) to match one PR pixel using an appropriate distance-weighted function. The ISCCP D1 product including clear-sky composites, surface and atmospheric parameters [Rossow and Duenas, 2004] is also used to provide primary threshold values for clear-sky identification (so-called cloud test) and to indicate atmospheric state required in the retrieval process.

[6] A preliminary examination has demonstrated that all PR-determined precipitating pixels pass the cloud test, implying precipitating pixels are completely equivalent to PCs, which is physically reasonable. Accordingly PCs are identified directly through PR measurements, while a cloud test technique relying on VIRS infrared channels at 3.75μm (CH3), 10.8μm (CH4) and 12.0μm (CH5) is then employed to distinguish N-PCs from clear-sky among the rest pixels. The cloud test firstly classifies as N-PCs those scenes with CH4 infrared brightness temperature (BT4) exceeding the uncertainty of the corresponding clear-sky value, within a 3-hours temporal range and 1° × 1° spatial domain given by ISCCP D1 products. And then cascading tests making use of interchannel brightness temperature differences (BTD), which are designed to detect low clouds or thin cirrus clouds [Baum et al., 1997] are implemented to fulfill the cloud test. It is noteworthy that acquired PCs do not include drizzle due to 17 dBz sensitivity limitation of TRMM PR [Kummerow et al., 1998]. Thus, PR-precipitating clouds (P-PCs) and PR-nonprecipitating clouds (PN-PCs) are specially termed in this study instead of PCs and N-PCs, respectively.

[7] A lookup table (LUT) method similar to the technique originally proposed by Nakajima and King [1990] is used to retrieve COT, where VIRS 0.6μm (CH1) and 1.6μm (CH2) are employed as the nonabsorptive and absorptive channel, respectively. The radiative transfer mode called SBDART [Ricchiazzi et al., 1998] is employed to establish the LUT. Taking into account the upward transmittance of surface longwave radiation that is dependent on COT and surface temperature, BT4 is further corrected to get an authentic CTT, which then is used to determine CTP according to the nearest atmospheric temperature profile given by ISCCP D1 products.

3. Results

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Data and Methodology
  5. 3. Results
  6. 4. Conclusions
  7. Acknowledgments
  8. References
  9. Supporting Information

[8] Consistent with the definition of ISCCP, CA is referred to as the ratio (hereafter expressed as a percentage) of the specific kind of cloudy pixels to the total observed pixels in a 2.5° × 2.5° grid box within three hours. The boreal summer averaged distributions of CA for PN-PCs and P-PCs are presented in Figure 1a and Figure 1b, respectively. Noting that the three scales in Figure 1 are different. Highly scattered and totally different spatial patterns are revealed, with the implication that areas with maximum amount of PN-PCs are not always accompanied by markedly high amount of P-PCs. Although both PN-PCs and P-PCs are plentiful in the Intertropical Convergence Zone (ITCZ) and have relatively similar distributions, the detailed patterns differ from each other. As shown in Figure 1b, there are a large amount of P-PCs located in ITCZ, Southern Pacific Convergence Zone (SPCZ), as well as monsoon regions in East Asia. In above regions, three maximum centers of P-PCs occur apparently over the Bay of Bengal, the West Pacific Warm Pool and the eastern equatorial Pacific, where the averaged CA of P-PCs exceeds 15. On the other hand, northern Indian Ocean, northeastern Pacific, southeastern Pacific and southeastern Atlantic are covered by a great number of PN-PCs with the averaged CA approaching 80 (Figure 1a). The West Pacific Warm Pool and the Bay of Bengal are shown to be two particular regions where NP-PCs and P-PCs are both abundant and the total CA could exceed 90. Moreover, it is noteworthy that the locations of maximum P-PCs and PN-PCs are not exactly consistent even in the Bay of Bengal, The former occurs much near the continent, which possibly implies a generating-precipitation effect of the coast topography on the southwest monsoon flows.

image

Figure 1. Global distribution of mean cloud amount of (a) PN-PCs, (b) P-PCs, and (c) the ratio of PN-PCs' amount to that of P-PCs, based on TRMM observations in three summers from 1998 to 2000.

Download figure to PowerPoint

[9] In particular, several near-coast oceanic areas, such as southeastern Atlantic, northeastern Pacific and southeastern Pacific, that are proved to be dominated by low-level maritime stratocumulus, have a large amount of PN-PCs but few P-PCs. As shown in Figure 1c, the ratio of PN-PCs' amount to that of P-PCs in a 2.5° × 2.5° common grid, which could be considered as a ratio of occurrence probability for the two categories of clouds, is as large as 500 in above regions and greatly higher than the rest areas. It seems that besides cloud-absent regions, areas predominated by low clouds are generally associated with infrequent P-PCs and correspondingly few surface rainfall. On the other hand, the least ratio is not less than 2 even in typical rainy areas, such as the Bay of Bengal, the Warm Pool, eastern tropical Pacific and northern South America, indicating that regional PN-PCs' amount is at least two times that of P-PCs throughout the tropics and subtropics.

[10] In spite of the noticeable discrepancy in the horizontal distribution of P-PCs and PN-PCs, the zonal mean CA tends to have relatively similar patterns especially if the CA magnitude itself is not considered (Figure 2). Both P-PCs and PN-PCs maximize near 10°N that is consistent with the position of ITCZ during boreal summer. Away from the peak at 10°N, both the two categories of clouds decrease rapidly northward to near 25°N over ocean and over land, southward to the equator over ocean and 10°S over land. Over ocean, a second peak of PN-PCs could be found at mid-latitude in southern hemisphere, while P-PCs in this zone are not as much as that in 10°N. Compared with P-PCs over ocean that shows evident dominance in summer hemisphere, PN-PCs are more evenly distributed across the equator. Over land, however, the latitudinal patterns of P-PCs and PN-PCs are quite similar to each other. Furthermore, taking into account the considerable difference of CA magnitude between P-PCs and PN-PCs, P-PCs exhibit a much more intense variability along the meridian with relative amplitude exceeding 100% both over ocean and over land, whereas the zonal mean CA of PN-PCs varies only within 10% and 40% over ocean and over land, respectively.

image

Figure 2. Zonal mean cloud amount of P-PCs (solid line) and PN-PCs (dashed line), over (left) ocean and (right) land. The bottom abscissa indicates PN-PCs' cloud amount and the upper one indicates P-PCs' cloud amount.

Download figure to PowerPoint

[11] Statistical results of CA, along with CTP and COT for P-PCs and PN-PCs are summarized in Table 1. According to the ISCCP cloud classification, 680 mb and 440 mb are used to further classify PN-PCs and P-PCs into low, middle and high subcategories. The total CA (PN-PCs plus P-PCs) averaged in boreal summer is 55.9 and 40.1, over ocean and land, respectively. Both values are a little lower than corresponding ISCCP CA (57.8 and 44.2, respectively), which may be largely caused by the lower resolution (8–10 km) available in ISCCP instruments as indicated by Minnis et al. [1999]. The P-PCs' amount, 4.2 (3.5) over ocean (land), accounts for 7.5% (8.7%) of the total CA. The mean contribution of P-PCs to the whole cloud coverage is less than 10%, consistent with the occurrence frequency of PCs estimated by Rossow and Schiffer [1999]. Since drizzle-accompanied clouds have been excluded from P-PCs, the derived CA of P-PCs should rather be regarded as a conservative estimation of PCs. If the global-averaged occurrence frequency of drizzle, about 0.5, suggested by Schumacher and Houze [2003] was introduced, the contribution of the ensemble PCs could approach as much as 15%. Among the three subcategories of P-PCs over ocean (land), high and middle ones account for 81.0% (85.7%) and 19.0% (14.3%), respectively, whereas the low P-PCs' amount is extremely small, suggesting that P-PCs are mostly associated with cloud top higher than 680 mb (about 3 km in tropics and subtropics) from the global perspective. On the contrary, low PN-PCs are dominant and account for near half the total PN-PCs, 45.5% over ocean and 40.7% over land, whereas there are only 28.6% (31.1%) and 25.9% (28.1%) high and middle PN-PCs over ocean (land), respectively.

Table 1. Statistics of Cloud Amount, Cloud Top Altitude, and Cloud Optical Thickness for P-PCs and PN-PCs, Averaged in Three Summers From 1998 to 2000 Over Global Tropics and Subtropicsa
 PN-PCsP-PCs
OceanLandOceanLand
  • a

    Low clouds, Pc > 680 mb; middle clouds, 680 mb > Pc > 440 mb; and high clouds, Pc < 440 mb.

Cloud amount:
   Low23.514.90.00.0
   Middle13.410.30.80.5
   High14.811.43.43.0
   Total51.736.64.23.5
Cloud top temperature, K268.6266.2231.5230.6
Cloud top pressure, mb635.7576.0290.9259.8
Cloud optical thickness6.29.164.168.7

[12] The averaged PN-PCs exhibit themselves as middle clouds with moderate COT, while P-PCs fall into high clouds with significantly large COT that is higher than 60 and approximately 10 times that of PN-PCs. There is a notable discrepancy of COT between PN-PCs over ocean and over land, with the former much lower than the latter. On the contrary, the ocean-land contrast of COT for P-PCs is insignificant. Since COT is primarily determined by CWP, the CWP of P-PCs seems likely independent of the surface situation. Moreover, it is noteworthy that the mean CTT of P-PCs, 231 K (230 K) over ocean (land), is a bit less than the threshold (235 K) used in GOES precipitation index (GPI) scheme.

[13] As proposed by Rossow and Schiffer [1999], nine ISCCP cloud types are designated based on CTP and COT, which are named according to historical cloud types, such as cumulus (Cu), stratocumulus (Sc), stratus (St), altocumulus (Ac), altostratus (As), nimbostratus (Ns), cirrus (Ci), cirrostratus (Cs) and deep convective clouds (Dc). By using the ISCCP classification scheme, an occurrence frequency in two-dimension space of CTP and COT, is examined for P-PCs and PN-PCs, respectively. Figure 3a indicates that most PN-PCs over ocean have COT less than 9.4. Cu (accounts for 24.2% of the total PN-PCs), Sc (20.8%), Ac (18.5%) and Ci (17.7%) are the four predominant PN-PCs types among the nine cloud types, occupying near 85% of the total PN-PCs. On the other hand, most P-PCs have high cloud top and moderate or large COT as shown in Figure 3b. Dc (accounts for 60.5% of the total P-PCs) is undoubtedly the first primary cloud type of P-PCs, showing agreement with the assumption of ISCCP classification scheme. Ns, another ISCCP-designated precipitating cloud type, however, occupies only about 4.0% of the total P-PCs while Cs (24.4%) is proved to be the second primary cloud type of P-PCs. To clarify the contribution of P-PCs for a given ISCCP cloud type that practically consists of both P-PCs and PN-PCs, the calculation indicates that Dc, Ns and Cs contain much more P-PCs than other cloud types. The P-PCs' fraction in these three cloud types is about 73%, 28% and 22%, respectively. The frequency pattern of P-PCs over land (Figure 3d) is approximately the same to that over ocean. Hence, it could be deduced that a considerable fraction of clouds with COT lower than that of ISCCP-defined Dc are actually associated with precipitation.

image

Figure 3. Frequency distributions in two-dimensional space of cloud top pressure and cloud optical thickness, for (a) PN-PCs and (b) P-PCs over ocean and for (c) PN-PCs and (d) P-PCs over land. Solid lines and the letter in each grid indicate the threshold and the cloud-type denomination suggested by Rossow and Schiffer [1999], respectively.

Download figure to PowerPoint

[14] Particularly, noting that the maxima are almost located at the dividing lines of COT, especially for those of P-PCs, where the occurrence is concentrated at 23.0. Since the boundaries for classifying ISCCP cloud types were designated so arbitrarily, the amount of unexpected “precipitating Cs” would decrease markedly if the boundaries were modified in some sort. On the other hand, these especial P-PCs with significantly high cloud top but moderate COT may represent multi-layer clouds that consist of lower convective or stratiform PCs and upper anvils.

4. Conclusions

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Data and Methodology
  5. 3. Results
  6. 4. Conclusions
  7. Acknowledgments
  8. References
  9. Supporting Information

[15] Based on TRMM measurements, CA, CTP and COT are investigated for P-PCs and PN-PCs. The global distribution of P-PCs and PN-PCs are totally different while zonal mean patterns are basically identical. Specifically, areas predominated by maritime stratocumulus are covered by a large amount of PN-PCs but extremely few P-PCs. Statistical results denote that total CA averaged across the tropics and subtropics in boreal summer is about 55.9 and 40.1, over ocean and land, respectively. The mean CA of P-PCs is near 4.0 and contributes less than 10% to the total CA. Low P-PCs that have cloud top lower than 680 mb are extremely rare, while low PN-PCs account for near half of total PN-PCs. Noting that drizzle-accompanied clouds have been excluded from P-PCs examined in this study, the contribution of actual PCs to total CA may exceed 15% if the missed-drizzle was taken into account. Likewise, the CA of actual PCs with low cloud top may increase to some extent as expected. The mean COT of P-PCs is revealed to be 64.1 (68.7) over ocean (land), approximately 10 times that of PN-PCs.

[16] According to ISCCP cloud types, Cu, Sc, Ac and Ci are four most fundamental cloud types of PN-PCs, whereas Dc and Cs, rather than Ns, are the first and second primary cloud type of P-PCs, respectively. In addition to Dc, a great amount of high clouds that have moderate COT, classified as Cs in ISCCP cloud types, are revealed to generate precipitation and account for total P-PCs by nearly 24.4%. Further statistics reveal that Dc, Ns as well as Cs involve considerable proportion of P-PCs, while other ISCCP cloud types almost consist of only PN-PCs. The unique precipitating Cs is probably correlated with the arbitrary boundaries of COT between ISCCP cloud types and is presumably multi-layer clouds that cannot be effectively detected through only visible and infrared signals. The characteristics of these special P-PCs with high cloud tops but moderate COT need to be further investigated in future.

Acknowledgments

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Data and Methodology
  5. 3. Results
  6. 4. Conclusions
  7. Acknowledgments
  8. References
  9. Supporting Information

[17] The authors greatly appreciate the anonymous reviewers for their valuable comments. TRMM data sets were provided by TRMM Science Data and Information System at the NASA Goddard Space Flight Center. This research has been jointly supported by NKBRPC grant 2004CB418304, Special Funds for Public Welfare of China grant GYHY-QX-2007, and NSFC grants (40730950, 40625014, 40605010, and 40675027).

References

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Data and Methodology
  5. 3. Results
  6. 4. Conclusions
  7. Acknowledgments
  8. References
  9. Supporting Information
  • Arkin, P. A., and P. Xie (1994), The global precipitation climatology project: First algorithm intercomparison project, Bull. Am. Meteorol. Soc., 75, 401419.
  • Baum, B., R. M. Welch, P. Minnis, L. L. Stowe, and J. A. J. Coakley (1997), Clouds and the Earth's Radiant Energy System (CERES) algorithm theoretical basis document, NASA RP, 1376.
  • Cess, R. D., et al. (1989), Interpretation of cloud-climate feedback as produced by 14 atmospheric general circulation models, Science, 245, 513516.
  • Fu, Y., and G. Liu (2007), Possible misidentification of rain type by TRMM PR over Tibetan Plateau, J. Appl. Meteorol., 46, 667672.
  • Fu, Y., G. Liu, G. Wu, R. Yu, Y. Xu, Y. Wang, R. Li, and Q. Liu (2006), Tower mast of precipitation over the central Tibetan Plateau summer, Geophys. Res. Lett., 33, L05802, doi:10.1029/2005GL024713.
  • Inoue, T., and K. Aonashi (2000), A comparison of cloud and rainfall information from instantaneous visible and infrared scanner and precipitation radar observations over a frontal zone in east Asia during June 1998, J. Appl. Meteorol., 39, 22922301.
  • Kummerow, C., W. Barnes, T. Kozu, J. Shiue, and J. Simpson (1998), The Tropical Rainfall Measuring Mission (TRMM) sensor package, J. Atmos. Oceanic Technol., 15, 809817.
  • Lin, B., and W. B. Rossow (1997), Precipitation water path and rainfall rate estimates for oceans using special sensor microwave imager and International Satellite Cloud Climatology Project data, J. Geophys. Res., 102, 93599374.
  • Liu, C. T., E. J. Zipser, and S. W. Nesbitt (2007), Global distribution of tropical deep convection: Different perspectives from TRMM infrared and radar data, J. Clim., 20, 489503.
  • Masunaga, H., T. S. L'Ecuer, and C. D. Kummerow (2005), Variability in the characteristics of precipitation systems in the tropical Pacific Part I: Spatial structure, J. Clim., 18, 823840.
  • Minnis, P., D. Young, B. Wielicki, P. W. Heck, X. Dong, L. L. Stowe, and R. M. Welch (1999), CERES cloud properties derived form multispectral VIRS data, paper presented at EUROPTO Conference on Satellite Remote Sensing of Clouds and the Atmosphere IV, SPIE The Int. Soc. for Opt. Eng., Florence, Italy.
  • Nakajima, T., and M. D. King (1990), Determination of theoretical thichness and effective particle radius of clouds from reflected solar radiation measurements Part I: Theory, J. Atmos. Sci., 47, 18781893.
  • Ramanathan, V. (1987), The role of Earth radiation budget studies in climate and general circulation research, J. Geophys. Res, 92, 40754095.
  • Ramanathan, V., R. D. Cess, E. F. Harrison, P. Minnis, B. R. Barkstrom, E. Ahmad, and D. Hartmann (1989), Cloud-radiative forcing and climate: Results from the Earth Radiation Budget Experiment, Science, 243, 5763.
  • Ricchiazzi, P., S. Yang, C. Gautier, and D. Sowle (1998), A research and teaching software tool for plane-parallel radiative transfer in the Earth's atmosphere, Bull. Am. Meteorol. Soc., 79, 21012114.
  • Rossow, W. B., and E. N. Duenas (2004), The International Satellite Cloud Climatology Project (ISCCP) Web site: An online resource for research, Bull. Am. Meteorol. Soc., 85, 167172.
  • Rossow, W. B., and R. A. Schiffer (1999), Advances in understanding clouds from ISCCP, Bull. Am. Meteorol. Soc., 80, 22612287.
  • Schiffer, R. A., and W. B. Rossow (1983), The International Satellite Cloud Climatology Project (ISCCP): The first project of the World Climate Research Program, Bull. Am. Meteorol. Soc., 64, 779784.
  • Schumacher, C., and R. A. Houze Jr. (2003), Stratiform rain in the tropics as seen by the TRMM precipitation radar, J. Clim., 16, 17391756.
  • Simpson, J., R. F. Adler, and G. R. North (1988), A proposed Tropical Rainfall Measuring Mission (TRMM) satellite, Bull. Am. Meteorol. Soc., 69, 278295.

Supporting Information

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Data and Methodology
  5. 3. Results
  6. 4. Conclusions
  7. Acknowledgments
  8. References
  9. Supporting Information
FilenameFormatSizeDescription
grl24351-sup-0001-t01.txtplain text document1KTab-delimited Table 1.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.