The Rain Shadow Effect

INTRODUCTION

I have always been fascinated with weather. One of the most interesting aspects of weather is rainfall, and its variance from one place to another. When mountains are nearby, the rainfall amounts can vary significantly within a small distance. One place this happens frequently is in the San Francisco bay area, especially in the southern end of the Santa Clara Valley.

There are two basic effects on precipitation caused by mountains. There is the "orographic" effect and the "rain shadow" effect (See Fig. 1). The orographic effect happens on the windward side of a mountain. The rainfall amounts increase dramatically as you move farther up the mountain on the windward side. The orographic effect is why Kentfield, in Marin County can receive 1300mm of rain per year while San Francisco only 500mm.

The other effect is the rain shadow effect. Since I live on the leeward side of a mountain, I am more interested in the rain shadow effect. The rain shadow effect is where precipitation amounts drop significantly on the leeward side of a mountain. This is why places like Arica, Chile average only .5mm of rain per year. The Andes Mountains, to the east, receive a lot of precipitation, but it is all gone by the time it gets to the Atacama Desert. The same thing happens in the Santa Clara Valley, although on a smaller scale.

I have been charting the weather at my house for about the last five years. I have also paid very close attention to weather statistics from the Internet, television and newspapers. Regarding the San Jose area, I have noticed that the southern end receives more rainfall than in the north. My house in the far southern end typically receives 30%-100% more rainfall than downtown San Jose. This is very interesting, since both places are in the same valley, and relatively close to each other. The Santa Cruz Mountains, to the south, can receive more than four times as much. Because of this, the South Bay is a perfect place to study the rain shadow effect.

One problem in studying the rain shadow effect is that information on this topic is very difficult to obtain. I have an entire bookshelf full of weather books, and none of them have more than a glossary definition. Most of them do not even mention the rain shadow effect. I searched a few libraries and book stores, but could find little more than a definition or brief explanation. I was also interested in detailed information on exactly how the mountains affect the rainfall in different parts of this valley. I could not find any detailed information on the subject.

Here are a few things that I would like to try to prove in my experiment.

 

HYPOTHESIS

 

Since the rain shadow effect is so complex, there are several parts to my hypothesis.

Because of wind, rainfall that is intensified by orographic lifting does not stop at the edge of a mountain. This is what causes there to be more rainfall at the southern end of the Santa Clara Valley than in the middle and northern end. (See Fig. 2)

As the intensified rain from the mountains passes into the valley, it loses it’s intensity at a constant rate. This rate is mainly dependent on temperature, upper-level wind speed, and the vertical height of the clouds.

There is a point where rainfall amounts no longer vary depending on the distance from the mountains (mountains in the direction the wind is coming from).

The distance from this point to the mountains depends on the temperature, wind speed and the vertical height of the clouds.

An equation could be made to estimate the ratio of the amount of rainfall before orographic lifting takes place to the amount of rainfall in the rain shadow.

 

OTHER QUESTIONS I WOULD LIKE TO ANSWER REGARDING RAIN

 

How far is rain blown on its way to the ground? (See Fig. 3)

Could the rain at Location 3 be coming from above the mountains?

 

MATERIALS AND METHODS

 

The hardest part of this experiment was obtaining the data. I needed very specific information. I needed several rain gauges, equally spaced, and oriented in the right direction. I looked though various books and the Internet, but could not find anything useful to me. Because detailed, highly specific information for such a small geographic location is nearly impossible to find, I had to retrieve the data myself.

I bought 7 high quality Sunbeam rain gauges. (Fig. 4)

I studied a topographical map and determined the ideal locations to place them. (See Fig. 4.5)

I used my altimeter, along with the topographic map to find the elevation at these different sites. (See Table 1)

I set up the 7 rain gauges. A few were placed in people’s backyards, one at an apartment complex, one at a church, one on top of a park sign, and one at my house. I had to ask for permission to set up most of them. (See Fig. 4.5 for a map of the locations)

A large storm arrived the evening after I set up the rain gauges.

After the rain had stopped, I checked all of the rain gauges, made sure none of them were stolen, recorded the data, and emptied them.

I recorded additional data from about 25 automated rain gauges located all over Santa Cruz and Santa Clara counties. Three of those were also used for the first part of the experiment.

I repeated this for the next two systems, which came soon after the first one.

RESULTS

 

The data was gathered from three storms (See Table 1). Two were very strong systems with a lot of wind and rain. The third had light winds and little rain. For each storm, I plotted the amount of rain in millimeters vs. the SSW to NNE distance in kilometers from the ridge of the 800 meter tall Sierra Azul Mountains.

(Refer to Table 1 and Fig. 5) For the first storm, the graph started downward at what looked like a constant slope. Then at about 17km it leveled off. There appeared to be a straight horizontal line from 17km to 39km. The mean slope of the graph from 0km to 17km was about 2.7mm(rain) per km(distance). The average temperature for the storm at Location 3, 6km NNE of the reference point, was about 13C. The peak winds during the storm at Location 3 were from the south at 15.2m/s.

The second storm came shortly after the first. The average temperature again was about 13C. The peak winds were from the south at 11.6m/s. The graph was somewhat different. The point where it leveled off was around 12km. The slope from 0km to 12km was about -3.6mm/km.

The third storm was different from the first two. The wind at the surface was calm for most of the time it was raining. The highest winds were only about 4m/s from the west. The average temperature during the time it was raining was 12C. There did not seem to be any kind of constant slope. Note - Rainfall amounts were so light that any pattern was most likely nothing more than an anomaly, and probably of no use to this experiment, but all data should at least be mentioned.

 

DISCUSSION

 

Precipitation is formed or intensified by air rising, cooling below the dew point, and condensing. The opposite is true as air descends. It is dried out and precipitation loses some of its intensity. Mountains force air to rise and fall. Simply stated, the rain shadow effect is caused by air warming up and losing its moisture after it passes over mountains.

Wind is the main variable. Without wind, there would be no rain shadow effect. There are two major effects of wind. First of all, the wind pushes the air against the mountains, causing it to rise and then fall on the other side, which means more precipitation in the mountains and less in the valleys. If that is the only effect the wind has, the rain patterns should be fairly uniform, and proportional with different storms of different magnitudes. There is, however, a second effect the wind has. Tall clouds, enhanced by orographic lifting, are pushed forward by the wind carrying with them some of the heavier rain that would normally fall in the mountains. Once they get over the mountains, they begin to dry out and lose some of their intensity. If all the other factors, such as wind speed, temperature, air pressure, air instability, and humidity are about the same, and they usually are in the valley during a winter storm, the rainfall should decrease in intensity at a constant rate. Newton’s First Law states that an object (such as a cloud) with no force acting on it moves with constant velocity. There is no force (besides a very small amount of wind resistance) acting on a cloud in the opposite direction the wind is blowing. Therefore, clouds will move at a constant velocity in the direction the wind is blowing. If precipitation is moving at a constant velocity and losing its intensity at a constant rate, the amount of rainfall from the mountains toward the valley will decrease at a constant rate. That is the basis for my main hypothesis, and it seems to be substantially reflected by the data.

There is also a limit to the rain intensity that is lost as it moves into the valley. In most storms, there is still plenty of moisture in the valley for rain, just not enough to maintain the added precipitation due to orographic lifting. Because of this, there should be very little variation in rainfall past the reach of the intensified rain forced in to the valley by wind. This hypothesis is also substantially reflected by the data.

(Please refer to Fig. 5 for the graphs) The first two storms showed a definite pattern, and seemed to confirm the hypotheses. The rainfall patterns both consisted of two main parts. The first part was sloped and the second was flat. The initial sloped line represents the transition from the mountains to the valley. That slope is dependent on the wind speed, and also on the height of the clouds. For higher winds, the line should have less of a slope because the effects are spread (blown) over a larger area. Weak winds should have steep slopes because the same effects are spread over a smaller area. When there is very light wind, the line should be very sloped, but since the rain shadow effect is minimal, the line would be very short. When there is no wind, there is no initial sloped line because rainfall should be equal on both sides.

Just as expected, the windier storm had a smaller slope than the storm with less wind. Also, taller clouds should travel farther before dissipating, so taller clouds would also produce a smaller slope. This was not much of a factor in these two particular storms, since the rainfall intensity (which should be proportional to the cloud height) was nearly equal in both systems.

Both of the initial lines are slightly curved. Since there are an infinite number of different wind speeds and cloud heights in a storm, there are an infinite number of slopes for the initial line. The addition of an infinite number of straight lines with different slopes produces a curve. Storms usually do not have the same wind speeds and cloud heights for equal time periods, so the shape of the curve is totally dependent on what percentage of the storm is at a certain intensity. More rain usually falls when there is more wind, so the curve should remain somewhat straight.

In the first storm. There is was steeper slope at the beginning, and it became less pronounced shortly thereafter. That was because that particular storm had two major parts. There was a period of heavy winds and heavy rain, which would produce a flat slope. There was also another long period of the storm with heavy rain and light winds. The steep slope which would represent that period, added with the flatter part, would produce an initially steep line which becomes less sloped where the effects of the lighter winds taper off. The rainfall vs. distance graph represents that very well.

The second storm had a fairly constant wind speed throughout the entire storm, which should produce a very straight graph. This proved to be what happened.

There is also a horizontal region on both graphs. The horizontal region represents the area unaffected by wind-blown rain and clouds. Since wind is no longer a major factor here, and the other variables (mainly temperature and humidity) are not significant enough, this area should be represented by a straight line. However, that does not mean that there is no rainfall variation past a certain point. It means that since the expected variables are too insignificant, the rainfall in this location is random and unpredictable. It depends on the particular storm, and the particular cells and bands that happen to develop. The rainfall in this area should not usually vary much. One exception might be during scattered thunderstorm activity where temperature, instability, and chance play a major role.

Just as expected, the graphs of both storms leveled off to nearly a straight horizontal line in the area farthest away from the mountains. The point at which the sloped part of the graph and the flat part meet should also be dependent on the wind speed and cloud heights, and it appears to be according to the graphs. The wind’s effects are carried farther with higher wind speeds, and with bigger clouds there is more to carry, so they should not dissipate as soon. The horizontal part of the graph should begin at a farther distance in a windier storm. Just as expected, the horizontal part of the graph of the windier storm began at a greater distance than in the other storm. So far, the first four parts to the hypothesis appear to have been sound assumptions.

It should be possible to make an equation to approximate the amount of rainfall at a given distance from the ridge of the mountains. However, I have no way of measuring some of the variables needed for such an equation. Some of the things I needed to know during each minute of the storm were the heights of the clouds, the upper-level wind speed, the lifted index (the stability of the air), the dew point at different altitudes, the temperature at different altitudes, and any other variables I may have overlooked. These values are impractical and difficult for anyone to measure accurately, even with the most sophisticated equipment. The rainfall variation in even the most contrasting storms is still relatively small compared to the amount of possible error. Even my best guess at these values would be so inaccurate that the derived equation would be practically useless. I can, however, make an equation to predict a less complex aspect of the rain shadow effect.

Although some is added by orographic lifting, the rainfall that was originally destined for the Santa Clara Valley also decreases after it passes over the mountains. The result is less rainfall in San Jose than on the other side of the Santa Cruz Mountains. Scott’s Valley is located on the other side of the mountains in the direction the rain usually comes from. Although Scott’s Valley was too far away for me to set up rain gauges, I did have access to their automated rainfall data via the Internet. I can use the data from the first two storms, along with the fact that there should be no difference in rainfall amounts when there is no wind and consequently no rain shadow effect, to make an equation to determine the ratio of rainfall between one side of the mountains and the rain shadow side. Since the rain shadow effect is almost entirely based on the wind speed, the wind speed is the only variable I need for a good approximation. By plotting the three points, I found the equation. The value for rain sub San Jose is at the Civic Center downtown. I chose this location because it is beyond the reach of wind-blown orographic clouds and rain. W is the peak wind speed at Location 3, 8m above the ground, 6km NNE of the Sierra Azul ridge. This is a poor value to use for wind speed, but it is the best I had. It should suffice, since it is close to being proportional to the average wind speeds aloft. It assumes that more rain falls when there is more wind, and the wind aloft is not extremely variable. Temperature should play a small role in this equation, but was not included because of insufficient data. Since air can hold more moisture when it is warmer, there would probably be a smaller difference between both sides when the temperature is higher, but that is a whole new experiment. Also, since San Jose and Scott’s Valley are 35 kilometers away from each other, a lot of random things can happen between the two places. Other variables besides wind and possibly temperature are too insignificant compared to the random variations that are possible.

It should be mentioned that not all storms are like the ones I studied. Some storms come from different directions. The purpose of this experiment was to investigate typical winter storms that come in from the west with southerly or south-southwesterly winds. These are by far the most common in this area. Although the numbers only apply to these types of storms, the concepts apply to every storm.

 

CONCLUSION

 

After doing this experiment, I am much more confident about my original theories. The results seemed to substantiate my hypotheses. The rain shadow equation seems to work, but I need more data and more opportunities to test it. Since I find the data so interesting, I will continue this experiment.


TABLES AND FIGURES
[ Rainshadow1.gif | Rainshadow2.gif | Rainshadow3.gif | Rainshadow4.gif | Rainshadow5.gif ]


*UPDATE: The Rain Shadow Effect won First Place at the Santa Clara Valley Science and Engineering Fair, and also First Place at the California State Science and Engineering Fair.

Please send mail to rainshadow@weatherpages.com with questions or comments about this project.


For information about the Olympic Rain Shadow in Washington, visit Olympic Rain Shadow Central.


THE RAIN SHADOW EFFECT

[ Home | Features | News | Search | Chat | Links
History | Webmaster | Contact Information ]

What's New | Links | Chat
Live Cams | Forecast Models
Glossary | Feedback
Home


 TABLE AND FIGURES

 . Rainshadow1.gif
 . Rainshadow2.gif
 . Rainshadow3.gif
 . Rainshadow4.gif
 . Rainshadow5.gif

Weather
Copyright 1998 Matt Haugland. All rights reserved.