Island-Induced Cloud Plumes Influence Tropical Atmospheric Measurements, Surface Radiation

McFarlane, S. A., U.S. Department of Energy

Radiation Processes

Radiative Processes

McFarlane, S.A., Long, C.N., and Flynn, D., Nauru Island Effect Study, Fourteenth ARM Science Team Meeting, March 22 to 26, 2004, Albuquerque, New Mexico.


Nauru Island, about 1,200 miles northeast of Papua New Guinea in the western South Pacific, is one of three instrumented island sites that comprise ARM's Tropical Western Pacific locale.


Nauru Island, about 1,200 miles northeast of Papua New Guinea in the western South Pacific, is one of three instrumented island sites that comprise ARM's Tropical Western Pacific locale.

A key objective of the U.S. Department of Energy's Atmospheric Radiation Measurement (ARM) Program is to identify interactions between the processes that determine the radiative properties of an atmospheric column, including clouds and the underlying surface. Early results of ARM's Nauru Island Effects Study (NIES) provide important new information on the effect of small tropical islands on local meteorology. Principal findings are that the small island of Nauru, one of three sites that make up ARM's Tropical Western Pacific locale, increases the amount of low cloud on the leeward side of the island from about 20% cloud cover to 30% cloud cover. The amount of middle and high cloud is not affected. This study has enabled ARM scientists to develop analysis techniques that can detect when an island influence is occurring and how surface measurements may be affected by the island-induced clouds. The study is also allowing scientists to quantify the extent of the island influence on cloud statistics and the impact on surface radiation. This information is important for the ARM Program, as well as for other programs that maintain small island measurement sites.

During a previous field experiment (referred to as Nauru99), low-level cloud plumes—shallow, cumulus clouds induced by the island and then conveyed downwind —were frequently observed on the western, or leeward, side of the island. Cloud plumes form when warm, moist air moves horizontally over the island and the diurnal heating of the island (relative to the surrounding ocean) induces low level convection and cloud formation. Although these clouds have little impact on the total radiation budget of the tropical Pacific region, they were identified as having the potential to bias cloud and radiation measurements taken at ARM's Nauru site. The NIES was carried out to identify the occurrence of island-induced clouds and to quantify their effect on ARM measurements.

The NIES consisted of two parts: the temporary installation of a moderately sophisticated set of instruments (including a laser ceilometer, multi-filter rotating shadowband radiometer, sky imager, and a suite of Eppley radiometers) at the Menen Hotel, followed by the installation of a permanent meteorological station with simple instrumentation nearby. During typical wind regimes, both installations were upwind of ARM's existing research site at Denig. The temporary instruments, installed in November 2001 and removed in May 2003, measured the existence of increased low-level cloudiness, and provided a means to compare long-term basic measurements to the existence of the island-induced cloud plume.

To identify plume-affected periods, the data were broken up into 1-hour periods between 0700 and 1700 local standard time. Only coincident data from both upwind and downwind island locations were included in the analysis. The average surface wind during the study period was used to determine the downwind location that might be affected by the island-induced clouds.

To quantify the plume effect on cloud frequency, the researchers examined the average plume—frequency of occurrence‐for periods that were plume affected and not plume affected at each instrument location. The average low cloud frequency at the location near Denig was 17% for non-plume affected periods and 31% for plume affected periods. At the NIES site, the average low cloud frequency was 21% during non-plume affected periods and 32% during plume affected periods. Thus, during the study period, island induced clouds increased the average low cloud frequency from roughly 20% to 30%. The cloud frequency at the downwind location was greater than or equal to that at the upwind location more than 75% of the time. Based on the results from the study, the absolute increase in low cloud frequency of occurrence at the effected site was 10%, and the average decrease in downwelling total shortwave was 50-65 W/m2.

Although the simple, permanent instrumentation is not sophisticated enough to unequivocally identify times when a cloud plume is occurring, it can be used to identify times when the plume is most likely to be affecting the surface measurements. Additionally, by having a measurement of downwelling radiation at a second location, the effect of the cloud plume on the measured radiation field at the upwind site can be quantified on a long-term basis.