Snowfall & Freezing Precip
Convective Weather
Turbulence
NWP
Water Resources
Land-Surface Processes
Remote Sensing
Precip Physics
Ceiling & Visibility
Surface Transportation Weather
Verification
Tech Transfer

I. Precipitation Physics

[Background]
[Feasibility Study for Rainfall Enhancement in the UAE]
[Hygroscopic Flare Characteristics]



1. Background

In many regions of the world, particularly in arid or semi-arid lands, traditional sources and supplies of ground water, rivers and reservoirs are either inadequate or under threat from ever increasing water demands. The Thirteenth World Meteorological Congress in May 1999 acknowledged this trend by citing the impending water stresses for two-thirds of the world's population by the year 2025. All fresh water, whether on the surface or underground, comes from the atmosphere in the form of precipitation, and to a lesser degree fog and dew. This has prompted atmospheric scientists to explore the possibility of augmenting water supplies by means of cloud seeding as one possible mitigation strategy among a multitude that could be considered (e.g., conservation, improved agricultural practices, etc).

Relatively recently, several researchers, including RAP scientists, presented reports on results from seeding experiments in South Africa, Mexico and Thailand using hygroscopic agents. These efforts included randomized cloud seeding experiments and limited physical studies, and provided strong statistical evidence suggesting increases in rainfall. The results suggested that the rainfall increases were due to seeded clouds persisting longer and producing rain over a larger area. In December of 1999, an international WMO Workshop on Hygroscopic Seeding further evaluated this emerging technology. The workshop participants concluded that while the recent experiments were quite encouraging, there were a number of cautionary points that needed to be kept in mind, including: the absence of an experiment showing an area-wide effect, the lack of physical understanding, and the use of radar alone to estimate rainfall. It was emphasized also that the results from the three experiments could not automatically be transferred to a new geographic area, since the background aerosol is thought to be very important in the process.

RAP scientists, in collaboration with university scientists, continued to investigate and develop the technology of hygroscopic seeding for rainfall enhancement, most recently in the United Arab Emirates (UAE) (see Figure 1). The primary goal is to offer advice on the efficacy of cloud seeding, while also advancing our understanding of the physical processes involved and training local scientists in the various aspects of meteorological research in general and cloud seeding technology in particular.

Figure 1. Map of the United Arab Emirates, located at the southern end of the Arabian Gulf (also known as the Persian Gulf).

2. Feasibility Study for Rainfall Enhancement in the UAE

In 2001, the UAE government initiated a RAP study to assess the efficacy and potential benefits of rainfall enhancement via hygroscopic seeding as a means to support freshwater resources. Several objectives were identified and intensive data collection was carried out during the winter and summer seasons of 2001 and 2002. The UAE program is designed to assess the infrastructure for measuring rainfall, to collate and understand historical hydrological data and identify relevant hydrological issues, to document the aerosol, trace gas, and cloud properties in the region over a range of meteorological scenarios and seasons, and to simulate natural and hypothesized seeded conditions via mesoscale, cloud-scale, and hydrologic numerical models.

RAP scientists Daniel Breed, Tara Jensen, and Vidal Salazar, under the direction of Roelof Bruintjes, conducted the field efforts in to collect the relevant data and to develop the necessary infrastructure for current and future studies on rainfall characteristics. David Yates directed the hydrological studies, and Janice Coen led the numerical modeling efforts working with researchers at the South Dakota School of Mines and Technology.

a. Cloud and Precipitation Studies

During the 2001 winter project, limited opportunities existed for cloud and rainfall studies. Less than ten days over the three-month period had sufficiently-developed clouds to investigate precipitation formation and seeding concepts. Assuming that the winter of 2001 had been unusually dry, additional data were collected in the winter of 2002. However, the results were similar, namely that less than a dozen days generated precipitating clouds and fewer suitable days for seeding existed in the winter season. Conditions illustrated in Figure 2 were typical in the winter.

 

Figure 2. Typical winter conditions under high pressure, with low level haze topped by boundary layer cumuli.

 

Although historical data indicated that precipitation fell predominately in the winter, albeit with wide variability, convective storms over the mountains in eastern UAE during the summer were also identified as potentially significant sources of rain. The amount of rain from these storms and their frequency of occurrence were poorly documented due to their remoteness. Therefore, field studies to investigate summer conditions were implemented. Data collected during the 2001 summer project showed that storms developed frequently over the Oman Mountains, but that many opportunities for airborne investigation were thwarted by airspace restrictions. Storms over the Oman Mountains east and southeast of Al Ain, in Omani airspace, consistently developed earlier and more often than over areas in UAE airspace to the north. Table 1 demonstrates this effect using radar data to determine the existence and initiation time of storms over the Oman Mountains in June and July of the field project period. During the period covered in Table 1 (46 days), storms formed over the Oman Mountains on 34 days. Storms formed in Omani airspace on thirty-three of those days, while storms formed in UAE airspace on only 16 days. Furthermore, on nearly all of those days, storms were initiated in Omani airspace an hour or more earlier. Storms such as these may be hydrologically important to the UAE.

 

Date Oman Airspace UAE Airspace Date Oman Airspace UAE Airspace
16-Jun     9-Jul 11:31  
17-Jun     10-Jul 9:21  
18-Jun     11-Jul 11:41  
18-Jun     11-Jul 11:41  
19-Jun     12-Jul 12:31  
20-Jun 9:11 9:11 13-Jul 16:01  
21-Jun 10:11 14:01 14-Jul 9:31 13:31
22-Jun     15-Jul 9:11  
23-Jun 10:51   16-Jul    
24-Jun 11:21   17-Jul    
25-Jun 8:01 9:11 18-Jul 9:01  
26-Jun 8:21 9:01 19-Jul 8:23 10:11
27-Jun 8:51 11:51 20-Jul 11:21  
28-Jun
 9:41   21-Jul    
29-Jun 10:11 11:21 22-Jul    
30-Jun
 9:21   23-Jul 13:21  
1-Jul 9:51   24-Jul 5:31 6:30
2-Jul     25-Jul 9:01 (none over mts.)
3-Jul     26-Jul 8:21 11:21
4-Jul 12:11   27-Jul 7:31 9:51
5-Jul     28-Jul 7:51 9:01
6-Jul 12:31   29-Jul 8:01 9:01
7-Jul 12:42   30-Jul 8:21 10:11
8-Jul 12:21   31-Jul 7:51 7:51

Table 1. Storm initiation time (UTC) in Oman airspace versus UAE airspace during June and July 2001

b. Aerosols and Trace Gases

Investigations of trace gases and their associated aerosols in the atmosphere are important for two reasons: their effect on climate and clouds, and the health and environmental impacts. Our focus in the UAE is the effect on clouds since aerosol particles affect the radiative properties and precipitation efficiency of clouds. Competing influences likely exist in the UAE between water-soluble aerosol particles, consisting mostly of sulfates derived from anthropogenic activities, and mineral dust, which includes such diverse compounds as quartz, clay, calcite, gypsum, haematite and others. However, when mineral dust exists in conjunction with other pollutants such as sulfates, the dust particles can become coated with the sulfate and therefore can be more active as CCN. Our primary focus during 2001 was to establish whether sulfates exist in concentrations abundant enough to affect the dust particles.

In collaboration with scientists from the University of Witwatersrand in Johannesburg, South Africa, flights were made over the entire UAE region to obtain spatial and vertical distributions of trace gases and aerosols. An example of the sulfur dioxide (SO2) distribution in winter is shown in Figure 3. It is easy to identify the main industrial activities that are associated with SO2 emissions. The three sites over the ocean represent out-gassing flares from major oil fields in the Arabian Gulf. The elevated concentrations inland seem to be associated with processing of oil in some way (e.g., oil refineries).

Figure 3. Average aircraft column SO2 concentrations over the UAE region during the 2001 winter field campaign.

 

Emissions of SO2 into the atmosphere result in the production of aerosols through oxidation. These aerosols are formed in the nucleation mode (0.01-0.1 ) and subsequently grow into the accumulation mode (0.1-3.0 ). Highest concentrations of these particles are likely to occur in the vicinity of the highest emissions of SO2 because of the relatively stagnant conditions (anticyclonic circulation) occurring in the UAE region in winter. When plotted in a similar fashion as Figure 3, aerosol concentrations measured by a condensation nucleus (CN) counter, encompassing both nucleation mode and accumulation mode particles, were indeed found to be highest at the identified SO2 sources. From such a composite map, mean vertical profiles of aerosols (measured by different instruments) can be constructed, such as that plotted in Figure 4 for the region near Dubai in northeast UAE. It shows that the top of the boundary layer during the winter is nominally around 1500 m and that concentrations of all aerosols decrease rapidly above the boundary layer. The large discrepancy between the CN and PCASP-measured concentrations indicate that the area is a large source of nucleation-mode aerosols. Also, above the boundary layer, it is likely that nucleation-mode aerosols are an important source of CCN. The data suggest that background levels of CCN can be enhanced by local pollution sources and therefore should result in higher cloud droplet concentrations, particularly when the clouds are coupled to the boundary layer. The potential interaction of sulfates with mineral dust remained a question after the 2001 studies, so measurements of the actual aerosol particles were planned for the 2002 airborne field campaigns. Scientists at Arizona State University provided an apparatus for collecting aerosols on TEM grids (filters for examination by electron microscopy), and are currently analyzing the results of dozens of samples. An example of the detail provided by this sampling technique is shown in Figure 5. Preliminary results suggest that sulfate-coated particles may not be as prevalent as hypothesized.

 

Figure 4. Concentrations of CN-measured aerosols (pink line), PCASP-measured aerosols (blue line), and CCN (yellow line) above the Dubai region of the UAE for January - March 2001.

 

Figure 5. Aerosol sample collected in 2002 near an oil field showing particles of NaCl (the square one near the center), calcium carbonate, calcium sulfate (probably gypsum) and common clay particles. Scale bar on the left is 1 micron.

 

3. Hygroscopic Flare Characteristics

A recent modeling study by Cooper et al. (1997) provided insights into the theory behind hygroscopic cloud seeding and gave guidance on optimizing hygroscopic flares. If the CCN introduced into the cloud from the seeding flare are larger in size than the natural CCN, the introduced CCN will activate preferentially over the natural CCN and change the character of the droplet size distribution to encourage coalescence and the formation of rain.

Several different manufacturers have started to make hygroscopic flares following the apparent success of hygroscopic seeding in South Africa and Mexico. It is important to evaluate the output particle spectra from these flares in order to be able to understand the effects they might have on the condensation/coalescence process and precipitation development in convective clouds. A test facility was designed to provide a reproducible environment for the combustion of flares and the measurement of the resultant particles. The facility was also designed to simulate the burning of flares on the wing of an aircraft (see Figure 6). An example of particle data from one type of flare is given in Figure 7. The time series plot at the top of the figure shows that the flare burned for about three minutes and that there was substantial variability in the output of larger particles during the burn. Although the mode of the average particle spectrum (middle panel) is around 0.3, a substantial portion of the aerosol volume resides is in particles greater than 1.5 (volume spectrum in lower panel).

Figure 6. Example of a burning hygroscopic flare over the UAE.

 

Improvements planned for the facility include optimizing the dilution process to best match the measuring capabilities of the aerosol instrumentation and adding some measure of the combustion temperature of the flares. Using the results of this facility, comparisons can be made, flare consistency can be checked, and ultimately, optimal flare characteristics can be measured and recommended.

 

(a)

(b)

(c)

 

Figure 7a, b and c. An example of particle data from a hygroscopic flare (manufactured by ICE, containing 65% potassium perchlorate). See text for more details.

 

[Top]