Because no one answered, I'll give it a shot as a layman.

Is the local wind-speed known when those satellite images were taken?

I mean, maybe because of rustling of the leaves, a percentage of the underside of them contributes to the 'spectrum'.

(Just a idea; don't chop my head off...)

Hi Lambert,

Yes indeed you are the first one to tackle this question. I must be honest, it is a question of which we do not know the answer yet. Be we have defined a research project in a remote sensing programme in my country, under which it should be possible, looking at different angles from an aircraft to a tree canopy to determine the measuring conditions under which it would be possible to make a distinction between the abaxial and adaxial sides of a a leaf.

Why would we like to know that yo ask ? Well with ground level leaf reflectance measurements we could already demonstrate that the ratio of abaxial to adaxial leaf reflectance is dependent on the impact of pollutants (fine dust, NOx, SO2 and soil pollution, and maybe other ones as well). Hence it is a nice biomonitoring approach in cities. For example in a street with lots of traffic this ratio changes as opposed to a street with only a small or insignificant traffic flux. We were able to demonstrate this in two cities in Europe, Gent and Valencia, hence a city in a Western European climate and one in a Mediterrranean climate. They both gave very positive results. Hence when we would be able to determine the measuring conditions at canopy level under which the difference between the reflectance of (mainly) the adaxial side, or (mainly) the abaxial side can be measured, we would able to map the impact of pollution (as a mytter of fact the integrated pollution impact on trees) in a lot more cities.

That might be pretty interesting from the epidemiological view as well, because we are also talking about the impact of fine dust on trees here as well.

So that's why I asked that question here above, which is a complicated one indeed. There is effectively the wind field which can play a role, but which can be avoided, by looking carefully at the meteo conditions during flights. Most importantly we think that the angular distribution of leaves on a tree plays a fundamental role in this venture. So the first focus will be on simulating trees with different leaf angle distributions (LAD's) and then check under which angles of view and azimuth ;-) we can detect a majority of adaxial versus abaxial sides of leaves and vice versa. There will evidently be a mix of both sides of the leaves being visible, but the trick is to find the conditions where these sides can be viewed in their most seggregated mode. And maybe some unmixing algoithm might help us to unmix the abaxial form the adaxial reflectances as well.

If we would succeed in separating abaxial form adaxial leaf sides at the canopy level, for certain frequently occuring tree species in a city, then we are in business for the biomonitoring of air and soil pollution in cities ising tree canopies.

I figger that might give the hot spots of air pollution as well as soil pollution in urban environments. With an airborne campaign of maybe three flights ( or 3 days) it would (might) be possible to completely map a city like Gent or Valencia (middle sized European cities) integrated pollution levels, and pinpoint the hot spots, to be able to study these more in detail and try to remediate, the high pollution levels whic sometimes are street canyons with lots of traffic as we saw in Valencia.

I hope we can shift the funding our way, to be able to perform this work

By the cay thanks Lambert for showing interest in the topic. It might be interesting in cancer research as well when we would overlay let's say a cancer incidence map with an integrated pollution map for a certain city in the Benelux. You never know what that would give. Maybe fireworks. Just imagine.

Cheers,

Frank

Thank you Frank for this interesting elaboration.

I thought you wanted to use satellite images for monitoring... I hope the airplanes you want to use do not contribute too much to the pollution you want to map!

I'm pretty sure that pollution maps and cancer incidence maps are highly correlated, but a causal correlation will be hard to prove.

Well, maybe not; can we get at data about where cancer patients live/lived and for how long? If one could find a high correlation at STREET level, then it is highly probable that pollution causes cancer.

I am sorry, but I have no time to hunt for those data. Maybe ask on this forum whether someone is interested in cooperating?

Keep up the good (fire)work!

Lambert

Ok Lambert,

I'll keep you posted anyhow.

Cheers,

Frank