Final Report


1/6/15 Final report to: City of Wichita Falls, Texas

Seeding Operations & Atmospheric Research (SOAR)

Executive Summary

Wichita Falls is under an ‘exceptional drought' condition and is implementing ‘Stage 5 Drought Catastrophe' restrictions. The city's main source of water comes from Lake Arrowhead, at 19.4 percent capacity and Lake Kickapoo, at 28.1 percent capacity. This drought condition prompted city officials to consider implementing an operational cloud seeding program targeting the watersheds.

Seeding Operations and Atmospheric Research (SOAR) was contracted to provide cloud seeding services. The objective of the Wichita Falls cloud seeding program is to increase rainfall in the Lake Kemp, Lake Kickapoo and Lake Arrowhead watersheds when rain-bearing clouds are known to have poor rainfall efficiency. The Wichita Falls project currently uses glaciogenic and hygroscopic seeding to improve the efficiency of the cold and warm rain process respectively.

Wichita Falls experiences a humid subtropical climate. In an average year, about 79 percent of the annual rainfall occurs during the warm season (March through October). The cloud seeding program operated cocontinuouslyor two operational periods. During the first period operations were conducted from 1 March to 30 June 2014. During the second period operations were conducted from 27 August to 3 October 2014. The first cloud seeding mission over the Wichita Falls target area was on 5th March and the last was on 18 September. The project performed a total of 38 missions and 61.9 total flight hours.

An independent analysis of the seeding operation was done by analysis of radar data for seeded clouds and non-seeded (control) clouds. The analysis concludes that seeding operations from March to September 2014 appeared to increase precipitation from the cloud bases as detected by radar. The total increase in rainfall was calculated to reach a total of 391,250 acre-feet of water over the Wichita Falls target area. Based on total values of rainfall increases of seeded clouds versus their matching non-seeded clouds over the Arrowhead and Kickapoo watersheds, the benefit/cost ratio for the cloud seeding operation is calculated at 1.58/1. This means that for every dollar spent by participating counties of the Wichita Falls program, the benefit is $1.58.

Operational cloud seeding for precipitation enhancement is conducted worldwide. Those conducting these operational efforts have weighed the evidence and concluded that the potential benefits from precipitation augmentation in their programs outweigh the risks and costs involved. Many of the current operational cloud seeding programs are being conducted in Texas where seven cloud seeding projects were in operation in 2013 (Figure 1). The history of and the rationale for the Texas operational cloud seeding programs have been addressed by Bomar et al. (1999). Most of the individuals involved in these efforts agree that the evaluation of seeding effectiveness in all the programs should have high priority.

These Texas programs are merely one of many manifestations that the quantity and quality of fresh water for human use is going to be a major issue during the 21st Century, especially in arid and semi-arid regions of the globe. Water in such areas is inherently scarce, and this shortage of water will be exacerbated by the continued migration of populations into these regions. The superposition of predicted changes in
climate and the deleterious effects of pollution on cloud forming processes and rainfall onto these dry regions could result in disastrous water shortages that preclude human existence. Within the past ten years most areas of the United States have faced frightening droughts only to have periods of wet weather remove the threat. The Southwestern U.S. is suffering still from major water shortages. These problems are occurring with increasing frequency and it is only a matter of time before catastrophic water shortages exact a terrible toll on human existence.

The potential responses to such a prospect are many and varied. One response would be to do nothing, allowing natural processes to run their course, teaching people to live within their water means or forcing them to vacate a region. Such a program of benign neglect costs virtually nothing to implement. The opposite course of action would be to develop programs of water conservation and enhancement to address the water needs of a region. No one opposes conservation, and most entities concerned about water have implemented water-conservation programs. In Texas, opinions vary on cloud seeding, depending on one's point of view. There are always people willing to invest in operational seeding, secure in their expectation that the benefits will exceed the costs involved.

This report provides a summary of the cloud seeding operations conducted in the Wichita Falls target area, describes the methodology used and presents an evaluation of the possible impacts of seeding operations.

1.1 Background
Wichita Falls is under an 'exceptional drought' condition and is conserving water in every way possible by implementing 'Stage 5 Drought Catastrophe' restrictions. These restrictions impose conservation practices on the use of city water. The city's main source of water comes from Lake Arrowhead, at 19.4 percent capacity and Lake Kickapoo, at 28.1 percent capacity (as of 2015/01/02). This drought condition prompted city officials to consider implementing an operational cloud seeding program targeting the Lake Kemp watershed, Lake Kickapoo watershed and Lake Arrowhead watershed. Seeding Operations and Atmospheric Research (SOAR) was contracted to provide cloud seeding services.

1.2 Drought assessment
The Wichita Falls area has begun its fifth consecutive year of drought. Figure 2 shows the current drought assessment for the state of Texas. Areas in north Texas, northwest Texas and the Texas panhandle are under exceptional drought (as of 2014/12/30). Figure 3 shows the fraction of drought coverage over the state of Texas. Extreme and exceptional drought hit the state in 2006 when at least 50 % of the state was under extreme drought. In 2011, more than 95 % of the state was in extreme drought with Wichita Falls receiving less than half of its annual rainfall. In the most recent drought analysis, less than 20 % of the state is in extreme drought, however the Wichita Falls area remains in exceptional drought indicating the area exhibits a slow drought recovery rate. Figure 4 shows the monthly precipitation plots compared to the historical average. Wichita Falls received 45 % of its annual rainfall in 2011, 68 % in 2012, 74 % in 2013 and 82 % in 2014. The rainfall has increased each year since the 2011 but the exceptional drought condition has persisted.

During the past four years, Wichita Falls has officially received 77.88 inches of rain since the drought began. Had normal rainfall occurred over the four-year period, the city would have received 115.68 inches (28.92 inches annual mean X 4 years), giving the city a 37.8-inch deficit. That's equivalent to the loss of one and a third year's worth of rain over the past four years. This rainfall deficit over the past four years has resulted in the city's two primary reservoirs dropping to 24 % of capacity, the city implementing stage 5 drought catastrophe water use restrictions, and the introduction of highly-filtered wastewater into the city's drinking water supply.

Figure 2: Texas drought conditions for 30 December 2014. (Source
Figure 3: Drought conditions over the state of Texas from 2000 to 2014. (Source
Figure 4: Monthly precipitation plots since the drought started in 2011. The green shows the total observed precipitation and the brown shows the historical average. Wichita Falls has had an average rainfall of 28.92 inches over the last 30 years. (Source
Figure 5: Monitored water supply for Wichita Falls area reservoirs (Arrowhead, Kemp and Kickapoo reservoirs). Source

There is often a lag time between when precipitation begins to rebound back up to historical norms and the refill rates of various reservoirs that were drawn down during the worst of the drought. Although the agricultural drought has seen occasional reprieves over the last four years, the hydroloigical drought has steadily woworsenedcross much of the region. Wichita Falls area reservoirs are at 24 % conservation capacity, a decline of 2.6 % from the previous year. Figure 5 shows the conservation storage capacity for Wichita Falls area reservoirs. Lake Arrowhead is seseverelympacted at 19.4 percent capacity and a decline of 7.8 % from the previous year. Since impoundment, Wichita Falls area lakes have dropped below historical low capacity records.

1.3 Climatology of Wichita Falls
In order to design a weather modification program that targets clouds that are most suitable for seeding, a climatology of precipitating clouds in the Wichita Falls area is necessary. This will ensure that the start and end dates of the seeding program and the type of clouds targeted will mamaximizeeeding effiency.

Main features
Wichita Falls experiences a humid subtropical climate (Köppen climate classification Cfa), with some of the highest summer daily maximum temperatures in the entire U.S. outside of the Desert Southwest. In an average year, about 79 percent of the annual rainfall occurs during the warm season (March through October). Monthly rainfall quantities ordinarily decline markedly in the colder months of the year, when frequent periods of cold, dry air from North American Polar Regions surge southward and cut off the supply of moisture from the Gulf of Mexico. The mean annual precipitation is 28.92 inches with the highest monthly mean rainfall of 4.15 inches occurring in June. Rainfall is typically greatest in early summer, with January, July, November and December being the driest months. Most of this precipitation is mainly convective during the warm season (May through October) and thus it is more localized and its spatial and temporal distribution as well as the intensity of rainfall is highly variable. Some stratiform type precipitation with more widespread and even rainfall amounts is typical in March and April.

It is widely accepted that the primary source of low-level moisture for the United States east of the continental divide is the Gulf of Mexico. However, Texas also receives moisture from the Eastern Pacific Ocean, moisture carried into Texas from the southwest by tropical continental airmasses. It is necessary to examine the synoptic conditions that initiate convection in the target area and to analyze the sources of moisture that fuels thunderstorms.

During the spring, convective initiation is caused by eastward moving cyclones. The topography of the region channels warm moist air from the Gulf of Mexico northward and cold dry arctic air from Canada southward. These two very different air masses, together with westerly downslope flow off the Rockies, have important influences on the convective mechanisms in the target area. The confluence of warm dry air off the southern Rockies and warm moist air from the Gulf of Mexico produces a strong westeast gradient in moisture called the dryline. The motion of the spring and summertime dryline in the absence of any large-scale weather systems usually exhibits a diurnal trend. In general, this trend is described as an easterly advance of the dryline during the daytime and a westerly retreat at night. It is known that convective development often is collocated with the dryline and such development may lead to thunderstorm initiation dependent upon various airmass properties in the vicinity of the dryline. Another important feature during this period is the cold front and how it combines with the dryline to act as a focus for convective development.

Early summer is characterized by a period of transition from the cold season circulation regime to the warm season regime. This is accompanied by a decrease in mid-latitude synoptic-scale transient activity over the continental United States as the extratropical storm track weakens and migrates poleward to a position near the Canadian border. The upper air becomes rather stagnant with a persistent pattern of a high-pressure ridge over the desert southwest. A prolonged east to west flow in the lower levels of the atmosphere increases the moisture from the eastern plains to the Continental Divide. Most of the convection remains confined to the mountainous areas of southwest Texas and southeastern New Mexico due to upslope low-level winds and upslope flow. Other areas rely on diurnal heating and humid conditions with some upper air feature support from disturbances rotating in the periphery of the high pressure for convection. When an upper level low pressure develops in the northern Pacific, the ridge breaks down and a trough pushes an occasional cool front, bringing some temporary relief from the heat in north central Texas. However, these fronts usually stall to the north of the region due to a strong southerly flow.

The latter part of the summer is characterized by the onset of the Mexican Monsoon in southern and southeastern New Mexico. The main synoptic feature is a persistent broad high-pressure ridge at the upper levels meandering over the southern plains and the desert southwest. This feature increases the convective inhibition due to synoptic scale subsidence over the region. The convective energy is weak and convection is hard to forecast due to the lack of a surface feature. The sea level pressure of the southwestern United States increases significantly and leads to the development of a thermally induced trough in the desert southwest. Surges of tropical continental air with convection along this trough are very common. Tropical maritime influences from the Gulf of Mexico may also be observed even though most of this moisture remains to the southeast. During this period the ridge over the western United States weakens as the monsoon high retreats southward and Mexican Monsoon precipitation diminishes.

1.3.2 Precipitation efficiency and cloud cover
Cloud seeding is only performed in clouds that have a low precipitation efficiency. Precipitation efficiency is the ratio of the surface rain rate to the sum of the surface evaporation and the vertically integrated horizontal and vertical vapor advection. For a thunderstorm with a typical life span of 25 minutes, only 10 % of the water vapor transported into the ththunderstorms eventually measured as surface rainfall. Other precipitating clouds that are associated with stratiform cloud systems can achieve a precipitation efficiency higher than 50 %. Therefore, convective clouds associated with thunderstorm development present the greatest opportunity for cloud seeding.

Thunderstorms are most likely to develop when less cloudiness is present. Figure 7 shows the median daily cloud cover in the WhWichitaalls area. The median cloud cover ranges from 22 % (mostly clear) to 52 % (partly cloudy). The sky is cloudiest on February 28 and clearest on July 26. The clearer part of the year begins around June 11. The cloudier part of the year begins around January 1. These data suggest that at the 75-percentile level, cloud cover starts to decrease in March and April with the lowest cloud cover in July. Cloud cover increases to high levels again in December.

Figure 7: The median daily cover (black line) with percentile bands (inner band from 40th to 60th percentile, outer band from 25th to 75th percentile). Data from (1974-2012).

1.3.3 Probability of precipitation
The probability that precipitation will be observed in Wichita Falls varies throughout the year. Figure 8 shows the fraction of days in which various types of precipitation are observed. Precipitation is most likely around May 20, occurring in 40 % of days. Precipitation is least likely around December 17, occurring in 27 % of days. Over the entire year, the most common form of precipitation are thunderstorms. Thunderstorms are the most intense precipitation type observed during 41 % of those days with precipitation. They are most likely around May 28, when it is observed during 26 % of all days. Thunderstorms are the most frequent precipitation type from mid-March to midOctober, and present the greatest opportunity for cloud seeding. Convective clouds start to develop in March when they are embedded in other cloud types. The frequency of thunderstorms increases in April and May when cloudiness decreases and thunderstorms are more isolated. In October, the frequency of thunderstorms decreases as the source of moisture from the Gulf of Mexico is cut off.

Figure 8: The fraction of days in which various types of precipitation are observed. If more than one type of precipitation is reported in a given day, the more intense precipitation is counted. For example, if light rain is observed in the same day as a thunderstorm, that day counts towards the thunderstorm totals. The order of intensity is from the top down in this graph, with the most intense at the bottom. Data from (1974-2012). 

2. The science of precipitation enhancement
2.1 The cloud seeding conceptual models
Different cloud microstructures require different treatment, if any, to improve their incloud coalescence. It has been shown in the previous section that the geographical region of the target area is influenced by diverse synoptic conditions during the growing season. Different thermodynamics produce different clouds with microphysical structures that affect the in-cloud coalescence in dissimilar ways.

The large and mesoscale dynamics determining the characteristics of the cloud systems down to the small scale microphysics determining the nucleation and growth characteristics of water droplets and ice particles all form part of the chain of events of precipitation development. In Figure 9, Bruintjes (1999) attempts to hypothesize on the physical path that leads to the initiation and development of precipitation. This involves various microphysical processes that proceed simultaneously but at different rates, with one path becoming more dominant because of its greater efficiency under given atmospheric conditions.

For the purpose of completion, for the benefit of the meteorologically uneducated reader, and at a great risk of oversimplification, it is useful to group precipitation mechanisms into those that involve the formation of ice particles and those that do not.

The cold-cloud mechanism postulates the nucleation of ice particles in supercooled clouds followed by their growth by vapor diffusion into snow particles. Under favorable conditions they may aggregate as snow or rime to form low-density graupel or snow pellets. This process is important in clouds of all types where temperatures are colder than about -15ºC, including the upper parts of cumulonimbus clouds.

The collision-coalescence process or the warm rain process occurs in relatively warm clouds with tops warmer than -15ºC and with bases warmer than +15ºC by the collision between water droplets. To produce the large amount of collisions required to form a raindrop that would eventually fall to the ground, some cloud droplets must be larger than others. Larger drops may form on larger condensation nuclei, such as salt particles, or through random collisions of droplets. Recent studies show that convective clouds that ingest polluted cloud condensation nuclei (CCN) suppress precipitation in the warm layer due to the large concentration of small droplets and will precipitate more slowly than a
similar cloud ingesting clean air, which forms small concentrations of larger droplets that coalesce faster into raindrops (Rosenfeld, 2001).

2.2 Cold cloud seeding
Since the discovery of glaciogenic materials more than 40 years ago, silver iodide has been the most widely used cloud seeding material in Texas. Silver iodide enhances the ice crystal concentration in clouds by either nucleating new crystals or freezing cloud droplets.

2.2.1 Static seeding concept
The static mode of cloud seeding is based on the concept that clouds are deficient in ice nuclei and therefore additions of silver iodide crystals that mimic the structure of ice should result in a more efficient precipitation producing cloud system. Cotton and Pielke (1995) suggest that seeding using this hypothesis is limited to:
a) clouds which are relatively cold-based and continental;
b) clouds having top temperatures in the range -10 to -25 ºC;
c) a time scale limited by the availability of significant supercooled water before depletion by entrainment and natural precipitation processes.

This hypothesis has been tested and scrutinized during the last decade in experiments with mixed results. Although there are constant indications that seeding can increase precipitation, a number of recent studies have questioned many of the positive results, weakening the scientific credibility of some of these experiments. As a result, there is some uncertainty as to the methodology of such a hypothesis.

2.2.2 Dynamic seeding concept
The concept of dynamic seeding is a physically plausible approach that offers an opportunity to increase rainfall by much larger amounts than the static concept. This concept is to seed supercooled clouds with large enough quantities of ice nuclei to cause glaciation of the cloud. Due to seeding, supercooled liquid water is converted into ice particles, releasing latent heat, increasing buoyancy, and thereby invigorating cloud updrafts. In favorable conditions, this will cause the cloud to grow larger, process more water vapor, and yield more precipitation (Bruintjes, 1999). The enhanced updraft may also promote the initiation of convection in the surroundings.

Rainfall increases of seeded clouds versus unseeded clouds are documented regularly in Texas. Although most of these evaluations show increases in rainfall mass estimated by radar, evidence on what the effect on area rainfall would be has not been documented.

2.2.3 Warm cloud seeding (Hygroscopic seeding)
The term "hygroscopic seeding" has been associated with warm cloud seeding. The objective is to enhance rainfall by promoting the coalescence process using hygroscopic salt nuclei generated by pyrotechnic flares or a fine spray of a highly concentrated salt solution. In addition, Cooper et al. (1997) illustrated that hygroscopic seeding might have a beneficial effect on precipitation development through either of two distinct mechanisms:
a) introduction of embryos on which raindrops form; or
b) broadening of the initial droplet size distribution resulting in acceleration of all stages of the coalescence process.

In 1990, G. Mather reported a case of inadvertent seeding of clouds by hygroscopic particles emitted from a Kraft paper mill in South Africa. This observation led to further cloud seeding experiments in South Africa (Mather et al., 1997), Mexico (Bruintjes,1999) and in Thailand (Silverman, 2002) with highly encouraging results.

2.3 Type of cloud seeding employed by SOAR in the Wichita Falls target area
The objective of a cloud seeding program is to increase rainfall when rain-bearing clouds are known to have poor rainfall efficiency. Texas projects use glaciogenic seeding to improve the efficiency of the cold rain process. The target of such seeding operations is the supercooled water which is found in the cold part of the cloud above the freezing level. 

2.3.1 The methodology of glaciogenic seeding
The agent used is silver iodide, which is released in clouds to empower the formation of ice aggregates. The maximum efficiency for aggregation occurs around –5 ºC. Seedable clouds must have top heights around this temperature but warmer than –15 ºC.

Typically, clouds are seeded at –5 ºC level and the silver iodide is released in the early stages of development and within the first half-lifetime. Dosages should reach the dynamic mode of seeding, around 100 ice-nuclei per liter of supercooled volume. Radar volume scans are used to measure the exact dosage required to reach the dynamic mode. Clouds that are seeded are very carefully chosen in that the in-cloud coalescence of the clouds is closely monitored using the Index of Coalescence Activity.

There are two methods of glaciogenic cloud seeding utilized by the SOAR program, 1) cloudbase seeding and 2) cloud top seeding.
     Glaciogenic seeding at base is accomplished by injection of 50 gram Silver Iodidepyrotechnic flares into cloud updrafts in excess of 200 feet per minute near the bases of clouds along predetermined tracks 10 to 30 miles upwind of the target area.
    Glaciogenic seeding at cloud top is accomplished by an aircraft flying just above the freezing level that drops up to two ejectable Silver Iodide pyrotechnic flares, each flare releasing 20 grams of Silver Iodide smoke, into the top of the growing cumulus cloud 10 to 30 miles upwind of the target area.

2.3.2 The methodology of hygroscopic seeding
In hygroscopic seeding the agent is salt (such as Calcium Chloride), which is released at cloud base to form large drops that increases the collision and coalescence process in the warm part of the cloud. This process is absent in continental convection. The hygroscopic salt enhances this collision and coalescence process and produces warm rain earlier in the cloud lifetime.

Clouds are seeded at temperatures warmer than 2 ºC in the early stages of development and within the cloud updraft. Dosages are dependent on the intensity of the updraft and the cloud base temperature. Radar scans and aircraft instruments are used to estimate updraft intensity. Higher doses of seeding agent are typically used when cloud base temperatures are warm and when updraft intensity is high.

Hygroscopic seeding at base is accomplished by injection of 500 gram Calcium Chloride pyrotechnic flares into cloud updrafts in excess of 200 feet per minute near the bases of clouds upwind of the target area.

2.3.3 Seeding material (pyrotechnics) used
The pyrotechnics used for seeding at cloud base and on top are manufactured by Ice Crystal Engineering (ICE). Reports from cloud chambers show that the glaciogenic flares are producing about 2 X 10^^13 ice-nuclei per liter at -5 ºC. Hygroscopic flares contain large cloud condensation nuclei (between 0.5 and 1 µm in diameter), for greater competition on the vapor, decreasing peak super saturation at cloud base, reducing cloud drop number concentration and increasing the mean cloud droplet diameter. This causes larger drops that coalesce faster into rain drop

2.4 Quantifying a seedable cloud
Clouds are considered to be seedable for increasing precipitation according to the static-mode seeding concept if 1) the collision-coalescence process is inefficient, 2) the rate of formation of supercooled condensate exceeds or is comparable to the rate of depletion of supercooled water, and 3) if there is sufficient time to grow seeding-induced precipitation particles that can reach the ground. These seedable criteria specified by the static-mode definition are difficult to apply practically because of the lack of quantification.

2.4.1 Extensive use of TITAN to quantify a seedable cloud
The TITAN (Thunderstorm, Identification, Tracking, Analysis and Nowcasting) software processes volume scan data from the Fredericks Oklahoma NEXRAD (Next-Generation radar) radar. The data allows analysis of different variables such as storm identification, location, area, volume, mass of precipitation, Vertically Integrated Liquid (VIL) as well as rates of variation of these parameters.

TITAN provides a tool for an appropriately trained meteorologist to quantify a seedable cloud appropriately. The meteorologist usually undergoes a series of decisions that may be best characterized as: Nowcasting, decision time, qualification, treatment, maintenance and termination.

1. Nowcasting is when the meteorologist is monitoring the atmospheric conditions desirable for deep convection and seedable clouds to form. This decision is taken after studying the meteorological model output that forecast the prevailing thermodynamic conditions. This process is usually a routine analysis of upper air conditions and surface conditions. Nowcasting of the possibility of thunderstorms follows and weather modification pilots are briefed accordingly.
2. Decision time is when the meteorologist decides to launch a seeding operation based on his/her observations of the current and forecast atmospheric conditions.This decision is usually taken after observing cloud echoes on TITAN and the echo development trend. Sometimes an operation is launched after watching clouds grow visually or by observing the surface temperature reach a threshold when convection is expected to initiate or intensify. For an operation with good timing, decision time should be preceded by qualification.
3. Qualification is when a cloud becomes seedable. This decision can be made visually by the pilot observing a cloud before it is detected by radar. Most frequently, a cloud is observed on radar before seeding occurs. In the Wichita Falls target area, a seedable cloud echo usually reaches a VIL (Vertical Integrated Liquid) of 10 kg/m2 and continues rising. The volume of the cloud echo should be in the order of 200 km3 with cloud tops above 8 km. The development trend of other clouds outside the target area is usually observed to determine the growth characteristics and the lifetime of the clouds. A short lifetime does not allow much opportunity for seeding. On TITAN, a seedable cloud usually shows a pocket of about 15% of the echo volume with a higher reflectivity at or slightly above 40 dBZ reflectivity at an altitude ranging from 6 to 10 km. This is characteristic of a cloud with weak coalescence and with a loading of supercooled liquid water above the freezing level early in its lifetime.
4. Treatment is the time after initial seeding. Occasionally, treatment may be preceded by qualification in isolated cases. In most cases, however, a cloud qualifies as seedable and the meteorologist instructs the pilot to start seeding. The seeding starts when the pilot encounters, locates or is directed to the updraft portion of the cloud where the agent is released. The updraft usually has to exceed 200 feet per minute.
5. Maintenance is when a constant rate of seeding is established with continued observations of growth in the echoing volume. During this period the cloud echo has not reached its half-life time. Careful analysis of the dynamic variables of the cloud 20 echo and their trend is necessary to define the half-life of the cloud. The pilot usually continues to experience updrafts and the meteorologist is able to locate areas of new growth within the cloud structure.
6. Termination is when seeding is stopped. A seeding operation is usually terminated either due to the absence of updrafts and/or due to the cloud echo exceeding its halflife time. 

3. Cloud seeding operations in Wichita Falls
3.1 Wichita Falls target area
The objective of the Wichita Falls cloud seeding program is to increase rainfall in the Lake Kemp, Lake Kickapoo and Lake Arrowhead watersheds when rain-bearing clouds are known to have poor rainfall efficiency. The Wichita Falls project currently uses glaciogenic and hygroscopic seeding to improve the efficiency of the cold and warm rain process respectively. The target of glaciogenic seeding operations is the supercooled water, which is found in the cold part of the cloud above the freezing level. Hygroscopic seeding targets the warm part of the cloud at temperatures warmer than freezing.

3.2 Cloud seeding equipment
3.2.1 NEXRAD radar
Precipitating clouds are monitored using the WSR-88D National Weather Service (NWS) Sband (10 cm) Next Generation Weather Radar (NEXRAD). NEXRAD radar does not attenuate in heavy rain, and they are operated continuously unless they are down for maintenance. NEXRAD data is available through SOAR. SOAR receives instantaneous reflectivity data from the NWS radar sites located in Frederick Oklahoma and Dyess Air Force Base to cover the Wichita Falls target area. NEXRAD data is run through TITAN as a graphic user interface.

3.2.2 The SOAR cloud seeding aircraft
The cloud seeding aircraft was a Cessna 340 equipped with hygroscopic and glaciogenic flares. The hygroscopic burn-in-place flares are installed on specially built racks on the trailing edge of the wing. The glaciogenic burn-in-place flares are also installed on the wing racks. Glaciogenic ejectable racks are installed on an ejectable rack on the belly of the aircraft. During seeding the pilot can chose between glaciogenic or hygroscopic seeding flares depending on the properties of the clouds.

3.3 Personnel on the SOAR team
SOAR employs personnel that are qualified for weather modification and atmospheric research. The SOAR program has been conducting weather modification operations and research since the 1990s. During the Wichita Falls cloud seeding program, SOAR personnel were available on a full-time basis for the duration of the project to ensure that all seeding opportunities were adequately targeted.

3.3.1 Project Manager
Mr. Gary Walker, the manager for SOAR, has been instrumental in the inception of the SOAR program and its management. Mr. Walker has extensive experience in weather modification piloting the SOAR aircraft. Gary Walker is responsible for the day-to-day management of the weather modification program in Wichita Falls.

3.3.2 Project Meteorologist/Radar Meteorologist
 Ms. Jennifer Wright is the Project Meteorologist/Radar Meteorologist for the Wichita Falls cloud seeding program and responsible for the day-to-day operations within the target area during the contract period. Ms. Wright is responsible for issuing weather forecasts and their dissemination, and to direct cloud seeding operations using TITAN to monitor seedable conditions and aircraft position.

3.3.3 Aircraft pilot-in-command
The SOAR cloud seeding aircraft is piloted by Mr. John Renoir. Mr. Renoir holds a Commercial Pilot License Multi Engine Land Instrument certification. Mr. Renoir has flown thousands of hours in cloud seeding operations and research projects. John Renoir has a lot of experience flying in Instrument Meteorological Conditions or in conditions when the pilot-in-command is flying in marginal weather conditions and flying the airplane solely on the aircraft instruments.

3.3.4 Aircraft mechanic 

The SOAR aircraft mechanic is Mr. Brian Moore. Mr. Moore has developed systems for cloud seeding operations on the SOAR aircraft and also maintains the cloud seeding aircraft. Mr. Moore also performed duties of aircraft pilot as required.24

3.4 Summary of operations
The cloud seeding program operacontinuouslyusly for two operational periods. During the first period operations were conducted from 1 March to 30 June 2014. During the second period operations were conducted from 27 August to 3 October 2014. The first cloud seeding mission over the Wichita Falls target area was on 5th March and the last was on 18 September. The project performed a total of 38 missions and 61.9 total flight hours. Clouds were seeded on 20 operational days (days with cloud seeding operations) and used a total of 410 flares (182 glaciogenic flares and 228 hygroscopic flares). Figure 15 shows the seeding locations in relation to the target area. Reconnaissance operations were conducted on 5 operational days when the aircraft was launched but the clouds were deemed not suitable for seeding.

4. Evaluation of cloud seeding in the Wichita Falls target area
Advances in remote sensing technology has established radar as the tool of choice for evaluation of cloud seeding programs versus the use of more conventional methods such as rain gauges. An independent evaluation by Active Influence and Scientific Management (AISM) using TITAN radar data has demonstrated that clouds seeded within the Wichita Falls target area are receptive to glaciogenic and hygroscopic seeding and responded positively producing additional rainfall (Ruiz, 2014).

4.1 Analysis of seeded clouds using TITAN
The TITAN evaluation is done using the non-randomized controlled study (NRS) method where the seeded cloud group is compared to the non-seeded cloud group. NRS is used in clinical trials where an experimental group is treated differently than a control group. One group is exposed to the conditions of the experiment and the other is not. TITAN calculations of differences (for any radar variable) are done by comparison between seeded and unseeded clouds (similar clouds or control clouds). The main assumption is that the control clouds model how the seeded clouds would have evolved in case no seeding took place. This assumption implies that the seeded clouds are expected to evolve in the same way as the control clouds, and that any difference in the evolution of the two groups is due to seeding. TITAN tracks the radar parameters for clouds in the seeded and control group and the analysis code operates on these parameters to look for differences between the two groups. In this case, the difference in precipitation mass is attributed to the seeding intervention.

The first cloud seeding mission was on 5th March and the last was on 18 September. Table 1 shows the dates and times of the seeding events that were analyzed in the TITAN evaluation. For the cases evaluated with TITAN, 39 seeded convective storms were identified: 9 small storms (precipitation mass smaller than 10,000 kilotons), 13 large storms (precipitation mass greater than 10,000 kilotons), and 17 type B storms (mature storms coming into the target area with age greater or equal to 1 hour). The corresponding estimate increases in precipitation mass using the TITAN evaluation

• 9 Small Clouds: ~ 10,627 acre-feet (an average layer of 0.39 in)
• 13 Large Clouds: ~ 295,219 acre-feet (an average layer of 0.82 in)
• 17 Type B Clouds: ~ 85,404 acre-feet (an average layer of 0.03 in)

The total increase in precipitation mass detected by the TITAN analysis was 391,250 acre-feet of water. This is an excellent result not only because of the apparent increases detected by TITAN, but because the cloud seeding operations were performed following the practice standards already established by the scientific weather modification community. Two seeding opportunities were detected (a storm over Wichita Falls County on June 25th and a storm over Cottle County on 28 August). Excellent timings (average ~ 96 %) and appropriate glaciogenic seeding doses (average ~ 40 ice-nuclei per liter) plus the support of intensive hygroscopic seeding led to the aforementioned increases.

Details about increases per county are offered in the technical evaluation report (i.e. see Ruiz, 2014). The relative increase for the whole target area has been estimated at 4.2 % of the precipitation but it reached the 5.8 % level for the center section of the target area.

5. Conclusions and recommendations
5.1 Conclusions
This report is comprehensive in describing the need for rainfall enhancement operations within the Wichita Falls target area. A careful examination of the climatology has shown that convective clouds form in the target area with a sufficient frequency to warrant a cloud seeding program. It has been demonstrated that such clouds are receptive to glaciogenic seeding and respond positively producing additional rainfall. The findings outlined in this report and the experience in the direction and management of the SOAR program has led to the following conclusions:

1. A climatoligical analysis indicates that the frequency of thunderstorms in the area and the associated weather patterns warrant the need for a cloud seeding program. The prevailing drought conditions limits the frequency of clouds, but the number of observed cloud systems that develop in the Wichita Falls area are big producers of rain and have responded positively to glaciogenic and hygroscopic seeding.
2. Clouds that do occur are receptive to glaciogenic seeding. An independent evaluation conducted by Ruiz (2014) shows that the relative increase in precipitation for the whole target area is estimated at 4.2 %. The analysis shows that seeding operations has increased the rainfall by 130 % in clouds that produce a precipitation mass less than 10,000 kilotons (small clouds), 113 % in clouds that produce a precipitation mass more than 10,000 kilotons (large clouds) and 10 % in clouds that are seeded after a lifetime of 1 hour (mature storevaluationvalutaion concludes that the total increase in rainfall equates to 391,250 acre-feet of water over the target area.
3. Using the analysis value for the total increase in rainfall from March to September 2014 of 391,250 acre-feet over the target area it is possible to calculate the cost per acre-foot of water produced by cloud seeding by considering the cost of the cloud seeding program and the cost of water to the City of Wichita Falls. By a conservative estimate it is assumed that 10% of the calculated total increase in rainfall (10% of 391,250 acre-feet) of 39,125 acre-feet is the total increase in rainfall into Lake Arrowhead, Lake Kickapoo and Lake Kemp reservoirs. Assuming that half of this amount in rainfall evaporates after reaching the three reservoirs, then the total increase in rainfall in the three reservoirs is 19,562 acre-feet. The Arrowhead reservoir covers 26% of the total targeted watershed and Kickapoo reservoir covers 9%. The corresponding rainwater increase over the Arrowhead and Kickapoo reservoir is 5,086 acre-feet 28 and 1,760 acre-feet respectively. If the cost of water per acre-foot to the City of Wichita Falls is assumed at $65 per acre-foot, then value of the increase in water in the Arrowhead and Kickapoo reservoirs is estimated at $444,990 ((5,086+1,760)*$65). The total cost for cloud seeding in the participating counties of the Wichita Falls program is $280,000 (March to October 2014). Based on total values of rainfall increases of seeded clouds versus their matching nonseeded clouds over the Arrowhead and Kickapoo watersheds, the benefit/cost ratio for the cloud seeding operation is calculated at 1.58/1 ($444,990/$280,000). This means that for every dollar spent by participating counties of the Wichita Falls program, the benefit is $1.58.

5.2 Recommendations
In view of the conclusions drawn above, the following recommendations are being made:
1. The results of the evaluation are a clear indication that the SOAR program is running a successful rainfall enhancement operation in Wichita Falls. It is recommended that cloud seeding operations resume in a timely manner to target convective clouds and thunderstorms in the Wichita Falls target area.
2. The climatology identifies time periods when precipitation is most intense in the area. Thunderstorms are the most frequent precipitation type from mid-March to mid-October, and present the greatest opportunity for cloud seeding. It is recommended that cloud seeding start on 1 April and end on 31 August with no interruption. The seeding period could be extended into September if the long term forecast is favorable. Historically rainfall in July and August decreases, however it is difficult to predict whether any particular year is a normal year. Seeding opportunities should not be missed.
3. It is apparent that one aircraft can effectively target the watershed however adding an additional aircraft will ensure that no seeding opportunities are missed. In the event that clouds develop at the western and eastern edge of the target area simultaneously, two aircraft can seed both areas separately and without interfering with each other. In the case that a single storm develops in the area, one aircraft could seed at cloud base and the other aircraft could seed at cloud top. This technique is used on other seeding projects and it is recommended for maximum seeding efficiency.


ASCE 39-03, 2003: Standard Practice for the Design and Operation of Hail Suppression Projects.
American Society of Civil Engineers, VA.

Arnett, D. S., 1980: Weather Modification by Cloud Seeding. Academic Press Inc., NY.

Bates, D. L., and A. Ruiz, 2002: Weather Modification Scientific Management in Texas: The extensive and
intensive uses of TITAN. Journal of Weather Modification: Vol. 34, pp. 104-110.

Bruintjes, T. R.,1999: A review of cloud seeding experiments to enhance precipitation and some new
prospects. Bulletin of the American Meteorological Society: Vol. 80, No. 5, pp. 805-820.

Cooper, A. C., R. T. Bruintjes, and G. K. Mather, 1997: Calculation Pertaining to Hygroscopic Seeding with
Flares. Journal of Applied Meteorology: Vol. 36, No. 3, pp. 1449-1469.

Cotton, W. R., and R. A. Pielke, 1995: Human Impacts on Weather and Climate. Cambridge University
Press, 288 pp.

Hobbs, P. V., 2001: Comments and Replies on "A Critical Assessment of Glaciogenic Seeding of
Convective Clouds for Rainfall Enhancement". Bulletin of the American Meteorological Society: Vol. 82, No.
12, pp. 2845-2846.

Mather, G. K., D. E. Terblanche, F. E. Steffens, and L. Fletcher, 1997: Results of the South African cloudseeding
experiments using hygroscopic flares. Journal of Applied Meteorology: Vol. 36, No. 11, pp. 1433-

National Academy of Science (NAS), 2003: Critical Issues in Weather Modification Research. Internet link:

Okabe, I.T., 1995: The North American monsoon. Ph.D. dissertation, University of British Columbia, 146

Rosenfeld, D., 2001: Satellite observations of the detrimental impact of smoke dust and air pollution on
precipitation. Institute of Earth Sciences. The Hebrew University of Jerusalem.

Rosenfeld, D., R. Lahav, A. Khain, and M. Pinsky, 2001: The role of sea-spray in cleansing air pollution
over ocean via cloud processes. Institute of Earth Sciences. The Hebrew University of Jerusalem.

Ruiz-Columbié, A., 2014: 2014 Evaluation Report Wichita Falls Weather Modification Program.

Silverman, B. A., 2001: A Critical Assessment of Glaciogenic Seeding of Convective Clouds for Rainfall
Enhancement. Bulletin of the American Meteorological Society: Vol. 82, No. 5, pp. 903-923.30

Silverman, B. A., 2001: Reply (to Comments and Replies on "A Critical Assessment of Glaciogenic Seeding
of Convective Clouds for Rainfall Enhancement"). Bulletin of the American Meteorological Society: Vol. 82,
No. 12, pp. 2847-2848.

Silverman, B. A., 2003: A Critical Assessment of Hygroscopic Seeding of Convective Clouds for Rainfall
Enhancement. Bulletin of the American Meteorological Society: Vol. 84, No. 9, pp. 1219-1230.

Strautins, A., T. Flanagan, and W. L. Woodley, 1999: Coalescence Activity in Texas Clouds: The Index of
Coalescence Activity and First-echo Tops. Journal of Weather Modification: Vol. 31, pp. 42-50.

Texas Water Development Board (TWDB), 2002: Water for Texas – 2002: Llano Estacado Regional Water
Plan. HDR Engineering, Inc.: pp. ES-20

Woodley, W. L., 2003: Assessment of the Effect of Cloud Seeding in the Operational Seeding Projects of
Texas During the Summer of 2002. Final Report on TDA Contract WM-02-014.

Woodley, W.L., R. Driori, D. Rosenfeld, S. Orr, and G. Bomar, 2001: Results of monthly and seasonal gauge vs.
radar rainfall comparisons in the Texas Panhandle. J. Wea. Rev, 33, 46-60.

World Meteorological Organization (WMO), 2001: WMO Statement on the status of weather modification.
Internet link: