I have an admission… I love weeds. Especially flowering ones.  Central British Columbia is cold in the winter, and often extremely dry in the summer.At certain times of the year, the weeds add a dash of colour to the landscape. In the spring, it is dandelions, and during the beginning of a dry August, it is hawkweed.

Hawkweed flowers look like small dandelions, and the seeds look similar, but not much else does. Instead of one flower per stalk, hawkweed flowers usually appear in clusters. The flowers and stalks are completely different, although the sap is still milky.  With the abundance of flowers right now, I decided to take some Ultraviolet photos.

Of course, before looking at the ultraviolet, My usual pattern is to beak the visible spectrum photo into the red, green, and blue channels. In the visible spectrum, hawkweed looks like this:

2018 07 11 12;25 _MG_8253

After reviewing my earlier posts. I’m making a change to how I present the differing channels. Instead of labeling the channels, I’ve coloured the borders. The red channel has a red border, green has a green border, and blue… I’m sure you get the idea.

Red and Green look very similar… in fact, they’re almost identical.

2018 07 11 12;25 _MG_8253 R2018 07 11 12;25 _MG_8253 G

The only real difference in a side by side comparison is that is the wires in the background have a slightly different shade. The foliage itself is so alike between the two channels that even comparing them by flicking between the two doesn’t turn up much- some of the background foliage (especially the wilting foliage) is slightly different, and the flower petals have a bit more contrast in the green channel.


The blue channel, on the other hand, is obviously different.

2018 07 11 12;25 _MG_8253 B

The foliage is much darker… that’s expected from the blue channel. Plants absorb a lot more light in the shorter wavelengths. The background rocks are brighter than the surrounding plants in this channel.

What isn’t as expected is the dark centre of the flowers. As this is a visible spectrum photo, This SHOULD be apparent in the visible spectrum, but I don’t notice anything even with the naked eye. My assumption is that the variation in the blue channel is so faint(the darker the channel, the less light reflected in that wavelength) that the red and green channels wash it out.

With my short pass filter( photographing blue, violet, and ultraviolet light), the pattern in the hawkweed flowers is much clearer.

2018 07 11 12;27 _MG_8271

It seems that the adrk centre is becoming more apparent as the wavlength shortens. The outside petals have a greater contrast to the dark centres in the shorter wavelength. This trend continues once the visible light is removed entirely.

2018 07 11 12;26 _MG_8256

The above ultraviolet photo( taken by combining my Schott BG40 and UG11 filters) shows that although most of the foliage is almost black, the flowers are bright… in fact, they are easily the brightest object in the picture. Stacking my lenses this way does seem to let in a small amount of infrared light. I can remove that by adding my short pass filter to the other two (at the expense of a much longer exposure). However, in this case, doing this doesn’t have much effect on the ultraviolet signature of the flowers:

2018 07 11 12;27 _MG_8267

The two photos are almost identical. Perhaps it is the infrared leakage, but the hawkweed flowers in the UG11+BG40 ‘double stacked’ photo have a faint yellow tinge, while the ‘triple stacked’ photo is mostly monochromatic. It’s all false colour anyways, as none of this should be visible to the naked eye, so the yellow tinge doesn’t mean much.

Given that the blue channel is so dark, I wondered how the dark centre would present itself if I were able to see deeper into the infrared range. I left the red and green channels unaltered, but swapped the blue channel for the UG11+BG40 ultraviolet photo.

The resulting photo:

2018 07 11 12;26 _MG_8256 fc

Fairly similar to the visible spectrum. The plants might be slightly darker, and the white wire is now yellow. The flowers look similar as well, but now the dark bullseye is apparent. The flowers look like they’re the same colour in the centre, but now become white near the edges.

I also tried a deeper false colour photo- in this one, I used the red channel of the visible spectrum, the short pass filter became the green channel, and the UG11+BG40 double stacked photo became the blue channel.

2018 07 11 12;26 _MG_8256 fc 2

The wind picked up while I took these photos; the most distant flower  moved too much, and I cropped it out entirely. The closer flowers moved slightly as well, so the photo isn’t a perfect match.

As the vegetation is so dark in the shorter wavelengths, red dominates this photo. The dark centre is clearly visible here, though.

And that’s what hawkweed looks like in the ultraviolet range. All good…. except…

There are 21 species of hawkweed in British Columbia. Eight are native to the province, and 13 are invasive. For the most part, they all look identical, and can cross pollinate with one another to create hybrid plants. Do they look different in ultraviolet? Good question. I have no idea. As I don’t know how to tell the species apart, I can’t see the UV signatures to see how they differ; as flowers each have a bit of individual variation even within a species, I would have to photograph a number of each species to see how they compare. Unless I can clearly distinguish the species, I can’t even try this.

There is one exception. Most of the hawkweeds have yellow flowers. One species does not- the flowers are a red/orange in the species Orange Hawkweed, even if the rest of the plant looks similar to the other varieties. This was the one plant I could compare.

The UV signature of the flowers in this species is verry different… in that there isn’t much of one. The flowers look more colourful in the visible spectrum:

2018 06 12 15;55 _MG_6836

…but are hard to distinguish from the rest of the foliage with the short pass filter:

2018 06 12 15;56 _MG_6842

… and in ultraviolet? They blend in so much that I couldn’t even focus the camera on the flower properly(usingUG11+ the Short Pass- I didn’t have my BG40 with me at the time).

2018 06 12 15;57 _MG_6850

… On the other hand, they are an invasive species. I don’t have to feel guilty about picking them, as long as I dispose of them properly. In fact, removing the plant is encouraged, and as the whole root system came up with the flower, I figured I removed a couple of them.

With the flowers picked, I have a better opportunity to investigate them. I used a pot full of potting soil, but no other plants.

2018 07 26 12;09 _MG_8570

Breaking orange hawkweed into the three channels shows a marked difference.

2018 07 26 12;09 _MG_8570 R2018 07 26 12;09 _MG_8570 G2018 07 26 12;09 _MG_8570 B

With the exception of the flower on the farthest right(with the dark centre), the orange hawkweed flowers all look the same in the red channel. The colour variation from the centre to the edge of the flower seems to mostly effect the green channel. The blue channel is uniformly dark… again, excepting the far right flower, which actually seems a bit brighter in the centre.

Is there any variation in the ultraviolet wavelengths? Well, when using the short pass filter…

2018 07 26 12;09 _MG_8575

… not really.  That right flower is a little different again, but most of them are just dark. With the BG40+UG11+Short Pass filters…

2018 07 26 12;13 _MG_8612

…. nothing.

However, there is a difference with just the UG11+BG40 filters:

2018 07 26 12;10 _MG_8586

The flowers are uniformly coloured, but do stand out from the dirt and foliage. They’re also blurry- I’ll get into that in a second.

As there is no difference between the flowers in the triple stacked photo, why do these ones look different? As I said, the UG11+BG40 filters let in a small amount of infrared light. I took a photo in the infrared range(with a Zomei 680 nm filter):2018 07 26 12;08 _MG_8565

In this wavelength, the flowers are almost white, contrasting to everything else being dark. They are much whiter than the stalks. What I think is happening is that while the infrared leakage isn’t usually a problem, the fact that the entire picture is so dark in ultraviolet, and the flowers are so bright in infrared, the IR contamination is brightening the flowers over a long exposure.

That would also explain the  blurred flowers. different wavelengths of light refract at different angles, and infrared wavelengths in the above photo are about twice as long as the ultraviolet wavelengths . By focusing my photo for ultraviolet light, the infrared light becomes unfocused; if I were to focus for infrared, ultraviolet would be out of focus. There isn’t an easy solution for this.

So, there is a significant difference between the two hawkweed varieties I can distinguish, which isn’t really a surprise. Maybe someday I’ll learn to distinguish the other 20 species… until then, though I’ll just enjoy the yellow flowers during the dry season of summer.


The above photos were taken with a modified Canon T3i camera