Melting-layer cloud observed during MR01-K05 cruise of Res/V Mirai

1. Introduction

Over the warm tropical oceans, Johnson et al. (1996) found prominent stable layers near 2 km (the stable layer of the trade winds), 5 km (near 0°C), and 15-16 km (near the tropopause) during the Tropical Ocean Global Atmosphere Coupled Ocean - Atmosphere Response Experiment (TOGA COARE). Johnson et al. (1999) also found that peaks in the vertical distributions of radar-echo tops exist in the vicinity of these stable layers heights.

The maxima of the cloud top frequency at specific layers (2 km, 5 km, and 15-16 km) over the warm tropical oceans are of universal importance to atmospheric energy budgets, because the reflection of solar radiation and longwave emission occur around the cloud-tops. Meanwhile, if cloud bases frequently exist at a specific level, the maximum of the cloud-base frequency must also play a central role in determining the tropical thermal structure through the heating by longwave flux from the lower level.

Therefore, in the present study, frequencies of the cloud base in the middle troposphere are examined from the direct observation, making use of 95GHz cloud radar and lidar with the dual wave-length (532 nm and 1064 nm).

2. Data and meteorological conditions during MR01-K05 Leg3-4 cruise

The data utilized in the present study is obtained over the tropical western Pacific by Res/V Mirai of Japan Marine Science and Technology Center (JAMSTEC). Stationary observation was conducted over the tropical western Pacific (around 1.85 N, 138 E) for the period of one month from November 9 to December 9, 2001 by Res/V Mirai. An active phase of convection of Madden-Julian Oscillation passed over the observational area in the last 10 days of the period. During the entire observational period, westerly were dominant in the lower troposphere. Accompanied with cloud activities, there were two peaks of strong westerly wind in the middle of November and early in December.

3. Data analysis and procedures

In the present analysis, clouds with a lot of falling condensate particles are excluded, and clouds with a little of (or no) falling condensate particles are only taken into account. The presence of a lot of falling condensate particles implies that diabatic heating and cooling vigorously occur within or below a cloud. The cloud-base height would be altered due to the diabatic motions, and the height is not the ultimate cloud-base height. Additionally, the heating by longwave flux would be much smaller than the heating or cooling through the condensation processes, and have little impact on determining the thermal structure of a cloud with a lot of falling condensate particles.

The level where cloud bases without falling condensate particles are frequently observed is examined, simultaneously using 95GHz cloud radar and lidar with the dual wave-length (532 nm and 1064 nm). The cloud radar and lidar observed to the vertical direction only. Cloud bases with no falling condensate particles are defined through the radar reflectivity and lidar backscattering, following three steps.

I. The first level where lidar returned energy exceeds certain threshold (Lc) in the vertical scan from the height of 2 km is defined as a provisional cloud base level (Zl). Lidar has the dual wave-length, and the lower level is selected as the Zl.

II. The first level where radar reflectivity exceeds certain threshold (Rc) in the vertical scan from the height of 2 km is defined as another provisional cloud base level (Zr).

III. The ultimate cloud base level (Zb) is defined by the Zl, when the height of the Zl is lower than that of the Zr, or when the Zl is defined although the Zr cannot be determined. Meanwhile, the ultimate cloud base level (Zb) is defined as indefinite, when the height of the Zr is lower than that of the Zl, or when the Zl cannot be determined. The Zb is also defined as indefinite, if even one instrument does not work,

4. Results

Figure 1 shows the vertical distributions of the cloud base frequency. In the present analysis, the Lc is varied from -4.25 to -5.25 by 0.25 (5 samples). The Rc is also changed from -10 to -35 by 5 (6 samples). Therefore, Fig. 1 contains 30 (5 x 6) samples all told. The error bar indicates the ranges within one standard deviation calculated from the 30 samples.

When the cloud base is deduced by lidar only, a prominent peak is found between the heights of 4.5-5 km (dashed line in Fig. 1). The melting level corresponds to the heights of 4.5 - 5 km, and the peak would result from the lidar bright band phenomenon. On the other hand, a broader peak is found in between the heights of 4.5 and 6 km with the use of the present analysis method (solid line in Fig. 1).