Chasing Clouds

I worked as a salaryman for 10 years after graduating from Keio, but I went to Australia on a business trip and liked it so much that I quit my job and started wandering around by car.

The scenery was flat and didn't change no matter how many hours I drove, but I suddenly realized that the presence of the landscape was completely different just by having clouds. From then on, the camera that had been pointing down gradually started pointing up, and before I knew it, I had become a photographer of clouds and the sky.

Since I was living in a tent, I quickly noticed even small changes in the sky. That was interesting. I would stay outside all day, just staring at the sky. I would continue those days of moving whenever there were no clouds for about 10 months.

So you were literally chasing clouds.

That's right. The first book I made from that was called "Sora no Iro" (Colors of the Sky). Thanks to its success, I was somehow able to make a living.

Once, while I was photographing a tree on a hill, the wind blew very hard, and clouds kept flowing behind it. As I was thinking about the composition through the viewfinder, I realized, "Wait, what I want to photograph isn't the tree, it's the background." From then on, my method changed to making the clouds the main subject and combining them with something else.

Even with the sky in the same place, the scenery is different after sunset compared to before dawn, for example, and the colors are inverted. It's fascinating how the scenery changes completely just by the expression of the sky changing. There are countless combinations of clouds and light, the color of the light, and things like evening colors or morning colors.

I think photographing the sky and clouds is very difficult in terms of how to crop the image and the timing. The look of the sky changes its expression in an instant.

When I'm driving, there are times when I think, "That cloud is interesting." Then, if I find something to combine it with, I stop. Usually, I'm excited at those times, so I take the photo in a rush, thinking, "Wow, that's amazing!"

From there, I set up a tripod and think about the composition, but usually, I only use the very first shot. In short, I don't take photos by thinking; I feel the vibration. I think the most important thing is my own surprise.

It doesn't matter if the horizon is crooked. If the feeling of wanting to take the photo is captured, it's better not to fix it. I often get scolded by other teachers for that, though (laughs).

Among photographers, there are those who decide on the composition precisely and take a calculated shot, and those who take photos intuitively.

When I first started photography, clouds were not recognized as a subject at all. Photo books of them were only like encyclopedias. It was an era when people said you would definitely fail if you entered them into a photo contest.

At that time, I changed my perspective a bit and took photos that weren't textbook-like at all—for example, a sky with a hole in it like a European religious painting, or shooting from underwater—and it actually appealed to the general public. Not so much to my colleagues, though (laughs).

Miyamoto-san, do you take photos from a meteorological perspective?

I like clouds, but I'm probably a different type than Kentaro Araki; I like physics more, and I conduct research on the reasons why typhoons form. However, I do occasionally take photos as an amateur.

The direct reason I aspired to study meteorology was that during my time in the Faculty of Science and Technology, while I was working a part-time job at a scuba diving shop on an island during summer vacation, it happened to be the year with the most typhoons hitting Japan ever. So almost all my work was like typhoon countermeasures, and I started from wondering, "Why do typhoons come?"

Somehow, it's exciting when a storm comes. The sky changes constantly, and when the evening light meets the rapidly changing sky, you can take incredible photos. That's why a clear day without a single cloud is the worst. I move on immediately (laughs).

Do you have a favorite type of cloud?

I suppose it's cumulonimbus clouds. I've also taken photos of them from above in an airplane.

Also, there are those island-like clouds lined up at the same height far away over the sea. I like those kinds of clouds and feel like photographing them.

Since I research ukiyo-e, I investigate various things depicted in them, and landscape is one of the very important themes in ukiyo-e.

If anything, the clouds in ukiyo-e are often depicted in quite non-realistic shapes rather than being drawn exactly as they are in nature.

The clouds are very stylized, aren't they?

Yes. The easiest example to understand is this ukiyo-e by Utagawa Hiroshige. The composition isn't realistic at all. It's a view of Mount Fuji from Suruga-cho, which is near the current Mitsukoshi Department Store slightly north of Nihonbashi. The thing in the middle is a cloud. The colors are strange, with yellow at the bottom, a bit of blue at the top, and the sky looking slightly red.

Clouds in Japanese painting primarily have a decorative purpose. Instead of drawing the shapes as they appear in reality, they are elements to make the screen gorgeous and elements to omit the space between the foreground and the distant scenery; these two are the basic ways clouds are used. During the Edo period, realistic expressions of clouds gradually appeared, but there are still many decorative clouds.

But clouds have been depicted since ancient times, haven't they?

Yes. Clouds themselves have been depicted since ancient times, even in the Heian period. However, in terms of usage, the religious element in Buddhist paintings is also quite strong. For example, Amida Nyorai is in the center, and various Bodhisattvas come from the Pure Land riding on clouds. This isn't unique to Japan; in European Christian religious paintings, clouds also serve as a connection to the world of God.

Therefore, clouds have not necessarily been depicted as realistic nature. I think it was from the 19th or 20th century that they began to be depicted as beautiful nature.

Before ukiyo-e, were there many Buddhist expressions?

There are others as well. For example, there are paintings called Rakuchu Rakugai-zu Byobu (Scenes in and around the Capital), which depict the scenery of Kyoto from the Sengoku period to the early Edo period, showing the streets of Kyoto from high above. In the depiction looking down over all of Kyoto, the golden things between the houses are all clouds. It's like golden clouds are floating within the city of Kyoto.

Creating a decorative screen rather than drawing realistically is one major characteristic of Japanese art.

Does that reflect the image people of that time had of clouds? I thought it seemed like something mysterious, or a story from the heavenly world.

There is certainly religious mystery, but the awareness of seeing clouds themselves as beautiful has existed since ancient times. In Japan, there is a sense of expressing the beauty of natural scenery as literature, such as composing waka poems about the shapes of clouds or how shadows are illuminated by clouds over the moon. In Sei Shonagon's "The Pillow Book," there are also passages that mention the beauty of clouds.

However, the sense of appreciating realistic clouds as a painting came closer to the modern era, and I think the "beauty of clouds" was quite a conceptual thing.

In European religious paintings, angels descend, and for example, in things painted on church murals, clouds always play a decorative role.

Later, in the 18th century, there was British landscape painting. Turner is famous, but a person named John Constable kept painting only the skies of Suffolk and was called the painter of the sky; his work is incredibly powerful.

That's right. As science developed in Europe in the 18th century, people like Constable and Turner in Britain saw the beauty of nature there. I think there was probably also a sense of seeing the world of God in nature, but the idea of capturing the beauty of nature as a painting emerged, and that later reached Japan as well.

In terms of the sense of seeing clouds as beautiful, Japan might have felt it earlier in the world of waka and such. On the other hand, in terms of preserving realistic clouds as shapes, I feel that Europe, especially Britain, found that aesthetic sense first.

However, in the case of clouds, the shape changes after a little time passes. So, in both European and Japanese paintings, they are used as parts in a sense. Like ornaments to enhance the main subject.

Nowadays, anyone can take photos, so there are many social media posts of sky photos. When I started, no one was taking photos of the sky. Now, because the sky is above everyone's head, it has become very popular as the easiest subject to photograph.

I think capturing clouds as a shape is very different depending on whether you draw them as a picture or take them as a photo. In "Suisaiga no Shiori" (A Guide to Watercolor Painting), a textbook from the Meiji era, it is written that drawing clouds in watercolor is difficult.

Mountains and trees don't move, so you can sketch them as they are, but clouds don't have the same shape for even a moment, so the shape in front of you keeps changing while you are trying to draw it.

On the other hand, regardless of the quality of the photo, people nowadays always carry a camera in their smartphones, so anyone can take a photo the moment they think the sky scenery is interesting. In terms of expressiveness toward clouds and the sky, it's completely different.

Nowadays, clouds are treated as one example of taking photos of flowers or familiar things. You can take them nearby without having to climb a mountain.

Also, videos of minor abnormal weather are often uploaded to social media. I think this is convenient because you can see the disaster situation, rain situation, and cloud situation at various points.

It's not limited to disasters, but for the past few years, there has been a lot of talk about whether data taken by users without meteorological knowledge through social media or apps can be used for weather forecasting.

In reality, however, there are issues with accuracy, so it's not being used yet, but if that can be cleared, it would be a huge amount of manpower.

On TV weather forecasts, there are segments where they introduce photos sent in by viewers. I've been sent photos by TV stations several times and asked, "Is this real?" Some of them are clearly processed, no matter how you look at it.

Distinguishing that must be difficult.

In my opinion, clouds are not something you wait for and photograph; they are something you encounter. So, I'm often asked, "Do you wait?" but I say, "I don't wait." I move and encounter them, and then I think 'Wow' and take the photo. It's a repetition of that. Furthermore, erasing power lines is out of the question.

It seems there are apps now that can change a cloudy day into a sunny day (laughs).

Anything goes. Even changing the weather (laughs).

HABU-san mentioned earlier that he likes cumulonimbus clouds, and cumulonimbus clouds are unique among clouds. The Earth's atmosphere makes it difficult for air to move up and down. Most clouds spread out horizontally. Rising vertically is a rare case and cannot happen without a special mechanism at work.

The key to that mechanism is that the bottom is warm and the air rises. That's why cumulonimbus clouds tend to form in the tropics or when the ground is warmed in summer.

How many meters tall is the largest cumulonimbus cloud?

The altitude goes up to the top of the troposphere (tropopause).

There are some that stop at the stratosphere and flow sideways. An "anvil cloud" is that kind of phenomenon, isn't it?

Exactly. When a growing cumulonimbus cloud reaches the top of the troposphere, it's about 15 kilometers, or 15,000 meters in total. Sometimes it's about 18,000 meters, but in Japan, it's about 12,000. When it gets there, it hits the top and has nowhere to go, so it takes the flat shape of an "anvil," which is why it's called an "anvil cloud."

Are there conditions for cumulonimbus clouds to form, such as high humidity?

Yes. High temperature and high humidity. The bottom is humid and keeps sucking up seawater and such.

During my Doctoral Programs, I spent about a month on the sea on an oceanographic research vessel, and at that time, it was 360 degrees of sea, literally like the sky on the horizon. Then cumulonimbus clouds would form all over the place every day.

Is the mechanism for the occurrence of cumulonimbus clouds and the occurrence of typhoons the same in some ways?

They are very similar. It's important for air to move upward in a typhoon as well, and since a typhoon is like a collection of many cumulonimbus clouds, if it's hot and humid overall on a scale large enough for many cumulonimbus clouds to form, it's a very favorable environment. After that, it's easy for it to grow if some kind of vortex, or wind, comes.

In the growth stage, a cumulonimbus cloud just keeps going up, aiming only for the top, but during that ascent, there are various processes within the cloud, and water droplets grow steadily. Eventually, those water droplets become large and heavy and fall, so rain falls to the ground. At that time, they drag the air down with them, so a downdraft, or downburst, occurs.

Looking at cumulonimbus clouds makes me feel excited. Their movement is fast, so they're interesting to watch, and I wonder how far they'll go.

Also, there are things called altocumulus, like sheep clouds, right? It's interesting when those are lined up and many small shadows are created by the light from above.

Listening to HABU-san, light and clouds are the keywords.

Sometimes when the sun sets, a streak of light shoots across the eastern sky on the opposite side. I wonder what that is all about. When streaks of light peek through gaps in the clouds, it's called "Angel's Ladder" (crepuscular rays), and those are quite cool and beautiful too.

Light is crucial to the beauty of clouds and the sky, isn't it?

That's true. I also love it when light is pouring across the sky, but you can see rain falling only in one specific part of the ocean.

We've had a lot of localized heavy rain lately. What is the mechanism behind that?

The details get a bit complicated, but roughly speaking, I think a major factor is the increasing amount of water vapor in the atmosphere.

Warmer air can hold more water vapor; you can imagine it like a larger sponge. Compared to the past, the size of the "sponge" has grown. When a trigger occurs and it starts to get squeezed, a huge amount of water comes out.

Does that mean the clouds themselves are getting bigger?

Probably so, but rather than just size, I think the clouds are becoming "stronger." The strength of a cloud can be expressed by the speed of the vertical flow inside it and the amount of water it holds. As global warming continues, this trend will likely persist, so I expect the frequency of heavy rain to keep increasing.

I have the impression that localized "guerrilla rainstorms" have increased compared to the past. Is there a specific reason why it's becoming so localized, rather than just the overall temperature rising?

Since the "sponge" is larger, the rain from a single cloud comes down in a sudden burst, but as clouds form, the air currents around them also change. When a cloud forms, it warms the surrounding air. It's the opposite of "uchimizu" (sprinkling water to cool down)—the opposite of heat of vaporization. When you feel a chill after a bath, it's because water droplets on your skin evaporate and cool you down, but here the opposite is happening. Water vapor that was evaporated turns into liquid, so it tries to warm the surroundings.

Warmed air is light, so it floats upward, affecting the surroundings and causing more clouds to gather. When two clouds come together, the rainfall naturally doubles. If they gather even more, they power up in a more complex way. We use the term "organization," and it may be that clouds are becoming easier to organize. Linear precipitation zones are an example of this.

So the overall movement of the atmosphere is becoming more and more active.

Since the way clouds form has become more intense, I think vertical movement has become more active. Conversely, without clouds, air hardly moves vertically at all. It just blows horizontally at the same altitude like a jet stream, but thanks to clouds, air moves up and down.

Is it like the atmospheric movement is being stirred up?

Exactly. You could say cumulonimbus clouds are a tool for stirring; their raison d'être is to stir the atmosphere. Because the air near the ground is too hot compared to the upper atmosphere, the system dislikes that temperature difference and wants to mix it. It's as if the clouds are there for that purpose.

It's often said that you can tell future weather movements by the shape of the clouds, but how much can you actually tell?

It has been said since ancient times that "with this cloud shape, it's about to rain," and it's actually quite accurate.

Weather forecasting is very precise now, but no one has yet verified exactly how accurate visual cloud shapes can be.

Even in the Edo period, there were books like "Tsuki Zukai" that explained what kind of weather to expect based on cloud shapes. I don't know how scientific they were.

On the other hand, even today, the current NHK morning drama "Okaeri Mone" features a protagonist who wants to be a weather forecaster. There are scenes where fishermen and foresters know from experience that the weather is about to turn bad. I find it fascinating and wonder exactly how much can be understood from cloud shapes.

For example, with a cumulonimbus cloud, it looks obviously problematic. The difficult part is that there are many cases where the weather turns bad later even if the clouds don't look that billowy.

In drama scenes, they'll say it's going to turn bad in the afternoon even when the weather hasn't broken at all (laughs). Is there a scientific basis for reading cloud shapes?

Yes. While it's based on empirical knowledge, many of those observations are scientifically sound. To change the subject slightly, modern weather forecasting is done by solving equations based on scientific knowledge using computers, and we've become able to predict clusters of cumulonimbus clouds with a certain degree of accuracy. The primary goal is to predict cumulonimbus-type clouds that lead directly to disasters.

On the other hand, it's difficult for current weather forecasts to predict other things, like the rows of small, fluffy clouds that Mr. HABU likes. In fact, there isn't much scientific evidence for exactly when the weather will break just because those appear; they are simply perceived as one type of omen. I think the presence of those types of clouds is part of a sequence where, if you follow the steps, cumulonimbus clouds form hours later and eventually bring rain.

One of the students in my lab is actually researching how people in the past thought about weather. They are currently focusing on the Muromachi period, but they're struggling because there isn't much information.

You hardly hear about it in the Muromachi period, but many agricultural books were produced in the Edo period. Naturally, weather and seasons were extremely important factors, so for people whose livelihood was farming and who had a deep connection with nature, awareness of the weather must have been high.

I haven't researched fishing much, but I don't have the impression that there are as many records compared to agricultural books.

Were cumulonimbus clouds and such often depicted in ukiyo-e?

I'm not sure if they can be called cumulonimbus, but for example, in a painting of Ryogoku Bridge over the Sumida River by Shotei Hokuju, a disciple of Katsushika Hokusai, there is a depiction of billowy clouds. This is from the 19th century, an era when copperplate engravings were being introduced from the West.

The awareness of depicting realistic landscapes was transmitted through art, so they imitated it by using perspective, and if you look closely, there are even shadows at the bridge. They were trying to imitate and incorporate Western expressions into ukiyo-e, where light hits an object and creates a shadow. This is an example of how new ways of drawing clouds were born that hadn't existed before.

So the expression became a bit more realistic?

Yes. I think the awareness of trying to draw clouds realistically gradually emerged toward the end of the Edo period.

Were Hokusai and Hiroshige the ones who depicted clouds and the sky most often in ukiyo-e?

Since Hiroshige and Hokusai produced many landscape works, the sky was naturally depicted often. However, in terms of unusual or somewhat realistic clouds, the ukiyo-e artist Kuniyoshi also drew unique clouds.

As we move into the Meiji era, clouds start to be depicted quite realistically. In the ukiyo-e prints of Kobayashi Kiyochika around Meiji 12 (1879), the landscape expression becomes quite realistic. I think whether or not an awareness of light emerges is very important in the expression of clouds.

Ukiyo-e lacks shading, and there was low awareness of expressing things like sunlight. That changed as they began to incorporate such expressions under European influence. As Mr. HABU mentioned, light is quite crucial in expressing clouds, isn't it?

Photography is interesting because, unlike painting, it captures a chance moment.

That's true. Cloud shapes change in an instant.

When you draw them in a painting, they can lose their realism or naturalness.

The fact that no two are ever the same is wonderful. Usually, even when traveling, I decide where to take photos in the evening, eat, and then shoot until sunset.

Is the evening when you can take the most beautiful photos?

The highlight is the evening. The sky changes moment by moment, and if a nice breeze is blowing at that time, the clouds drift into various shapes, and the way the light hits them changes every time. In the end, it's about the combination—like the sunset reflecting beautifully on mackerel clouds.

Have you ever seen a green cloud?

I've heard of them, but I've never seen one. They appear above the horizon, right?

What kind of phenomenon is that?

It's called a green flash. The color of the sky changes because sunlight originally contains rainbow components, and the color is determined by how much atmosphere the sunlight has passed through to reach that point. During the day, it's blue because it's directly overhead and the distance is short, but in the evening or morning, the distance becomes longer, so the blue disappears and red remains—but there should be green in between.

It's a very rare phenomenon where a bit of green appears for just a split second.

I haven't seen that, but in the flat expanses of Western Australia, I often see clouds I've never seen before. Like when it's not raining only over the road, and it looks like a picture frame.

I take photos of clouds in Japan too, but in my neighborhood, there are too many obstacles like power lines. Australia is great because there's nothing. On a cloudless day, if you look 360 degrees, you can see somewhere that looks like it has a slight haze. I've driven 400 kilometers toward a spot like that, thinking there might be clouds there. I've been doing things like that for about 30 years. It's fun, chasing clouds.

A wide sky is wonderful. When I was a student, I studied abroad in Oklahoma in the US for a year. It's famous for tornadoes, and it really felt like there was nothing but the sky and the plains.

There are people called "tornado chasers" who chase tornadoes for business; they wait in their cars, and when a tornado is about to appear, they chase it, film it, and make a living by selling the footage to TV stations.

For some reason, those people always manage to find the tornadoes.

Do they predict the tornadoes and then chase them?

They might be looking at radar, but the data isn't that precise, so maybe it's intuition. Tornadoes almost certainly form under a "parent cloud," which is an even more powerful version of a cumulonimbus cloud, so the challenge is whether you can capture that parent cloud.

Speaking of Australia, do you know about the Morning Glory?

Yes. It's like a white mist in a roll shape that spreads out horizontally. I think I actually drove into one once.

Is that so? (laughs)

When you're driving, suddenly your visibility drops to about 50 centimeters.

Is it a downpour of rain?

No, it's not raining. You just suddenly enter a very thick fog. Unless you're quite far away, you can't tell how big it is.

That's dangerous in a car. Is the Morning Glory a cloud?

I think it's like a mist. If you were to classify it, would it be a cloud?

Even when we say "clouds," they come in many sizes. Clouds are actually made of water droplets or ice particles, but newly formed clouds are about 0.01 mm or 10 microns—thinner than a human hair and incredibly small.

However, when they become falling cloud particles or raindrops, they grow to several millimeters, which is about 100 times larger than a newly formed cloud particle. So, even in the same cumulonimbus cloud, a part that was the size of a marble when it was first formed becomes the size of a giant ball used in a ball-rolling race by the time it falls.

The particles in a Morning Glory are probably about the size of a newly formed cloud. About the same as mist. Also, the size of the sand particles kicked up in a dust storm isn't that different. I imagine that once you're inside, you can't see anything at all, so it would be so intense that you wouldn't know if it was mist or a sandstorm.

Well, there really are many surprising natural phenomena that one has never experienced before. Truly.

Are there more students now who are interested in meteorology or want to become weather forecasters?

Yes. The number is increasing every year, and many people are motivated by environmental issues. Others are interested in renewable energy and disaster prevention. My lab currently has about 40 people, including graduate students.

We set aside time for "sky observation" at the start of our seminars. Since we are online now, I have them go outside their homes and report on the sky conditions in their respective areas.

For weather forecasters, taking a photo inevitably becomes a form of recording. I recently had a conversation with Mai Kikuchi, a Keio graduate and weather forecaster, and she asked me, "How can I take photos like this?" I told her not to think when she takes them. If you think, the extra thoughts you had will inevitably show up in the photo. I told her to just take the photo if she thinks it's beautiful.

When we look at clouds, we first think about which of the ten cloud genera they belong to. We also focus on academic aspects, such as how much cloud cover there is across the entire sky.

But it seems like a good idea to have them capture and report things they find "beautiful" at first glance. It would also help cultivate their sensory perception.

The Meiji-era textbook I mentioned earlier, "Suisaiga no Shiori" (Guide to Watercolor Painting), also features the concept of the ten cloud genera.

Is that so?

It was published in 1901, and it says things like, "Cloud shapes can be divided into ten types, and for this type of cloud shape, use this color." Since they had to sketch what they saw, I suppose it was necessary to record the shapes using not just sensory perception but also academic and conceptual knowledge.

After all, artistic beauty and scientific aspects are never two separate things; they have commonalities, and I believe they have a deep relationship with each other.

I've been taking photos of clouds for over 30 years, and I never get tired of it. No matter how many I take, it's never the same, and there are always new discoveries. I feel like I can keep taking them for the rest of my life.

Clouds might truly be the place where science and art intersect. Science and art are often thought of as opposites, but I think it's better to mix them, so I intend to stay interested in many different things.