Friday, December 19, 2008

Data management

After a successful field (and holiday) season that had started already in early may, I returned to my office in mid-september. Then I started to realize that I had several gigabytes of raw data that needed to be analyzed: hemispherical images, wide angle aerial images, GPS and laser measured positions, stand- and plotwise forest inventory data, crown radius measurements, individual tree measurements… I was of course aware that it would take several weeks to go through it all, but I was not quite able to see that it would last until today, 19. December, to finally have it all neatly in (less than five) spreadsheets! So the autumn has been mostly copying data from one place to another, running some analysis, and then copying again. Of course there were also some more pleasant (coding) phases when a feeling of creating something new occasionally emerged, but what I’m likely to remember afterwards are the endless rows of data – and the particularly frustrating task of bug fixing! But now it’s finished, and next year there will be plenty of “real” research to be done.

Merry Christmas and happy New Year!

Height modeling

Tree positioning

Crown mapping

Image thresholding

Monday, November 3, 2008

The joy and despair of proposals

I went through my files today to see how many of our group's research proposals during 2002-2008 have received funding. 10 proposals (both small and large) have been funded, 23 have failed. By looking at those statistics I feel a bit desperate. And just like most of my colleagues, a certain fatigue overcomes me when I think about the next round of funding applications we need to write, and the long weeks it takes. I already have a premonition about the next round of disappointing news we have to hear.

Two years ago I felt so desperate and nervous that I decided to quit science. All the funding proposals by our group had been rejected for quite a while. A close friend of mine, an artist, was having similar feelings and difficulties in financing her artwork. We had the perfect solution to our problem: we would both become stewardesses! Finnair was launching a new stewardess training course (just three months) right at that time, so we both applied. I pictured myself on long intercontinental flights and strolling around foreign cities. It seemed like a great idea!

For some reason, a girl with a PhD in forestry did not qualify for even the first interviews. I hadn't realized that getting a job, whether scientific or not, would be so difficult with the type of education I had. I had been brainwashed to believe that studying forestry would be a good, all-around education. My artist friend was invited for the Finnair interviews, but did not qualify for the job either. So, there we were once again, trying to figure out what to do in the future.

Two weeks later things changed. I received the news that I had been granted a three-year postdoc fellowship. It was pure euphory for several days! I was reliving the moments I had had four years earlier when I received a full scholarship to work on my PhD.

Why am I writing about this today? Two days ago I had The News again. I have received my first big project as PI, for three years.

And yes, I was too excited to sleep the night after hearing the news. The joy phase is on again!

Friday, September 12, 2008

SNS seminar in Iceland

Most Finns want to travel to hot and sunny regions whenever possible, but I like cool places even more. So when the opportunity to participate SNS forest inventory seminar in Egilsstaðir, East Iceland, emerged in last winter, there was no reason to hesitate. Even though our Joensuu university delegation’s pre-journey feelings suggested that this trip should probably be classified as “scientific tourism”, the official part of the programme included many interesting presentations about NFI and laser scanning topics. Also the country reports gave a clear view of the related research in the Nordic and Baltic countries. Highlight of my own presentation was the international revelation of the crown relascope, a new instrument for canopy cover measurements. Still, when the last “never-ending” session was finally finished, everyone was quite happy to get out of the auditorium to really see Iceland.

An old Icelandic joke reveals an answer to the question ”What do you do if you get lost in an Icelandic forest? Just stand up”. It soon became clear that the modern Icelandic foresters don’t much like this joke. After visiting several forests near Egilsstaðir and Akureyri it became clear that they have a good reason to be proud of their success in increasing forest cover in a rather hostile environment. When the first tree in the country reached 20 meters height, the prime minister solemnly placed a memorial plate on the site of the historical event. Although we heard that some Icelanders are afraid that afforestation will destroy their open sceneries and make the country “look like Finland”, they are still very attracted to camping in forests by the lakes and rivers. Controversial?

Finnish larch going strong in the middle of the North-Atlantic

Not to be afforested in the near future

Wednesday, September 10, 2008

A change of environment

Last week I got my first touch to field work in a completely new environment - eucalyptus plantings and South Australian native vegetation remnants - during a day trip to a test site where several hyperspectral (and LiDAR) remote sensing projects are carried out by my new colleagues at the University of Adelaide.

Before the trip, I had a glimpse at some hyperspectral images from the test site and browsed through some forest inventory data. However, I still couldn't imagine what the area would look like in real life.

Very different from what I was accustomed to, that's for sure. There were about 30 different tree species in a small area (I didn't know any of them), mallees, the ground was bare (with patches of green and yellow grasses). One thing is in common with coniferous canopies: the grouped structure of many of the eucalyptus crowns. I guess it's a good starting point to find some common factors in a new environment - psychologically, at least!

So, why the field trip? During the next few months, I will be collaborating with the local researchers and testing some of the approaches (such as the 'p-theory' for modeling canopy reflectance) that I have been working with in the boreal forests. This time only the environment will be very different -- hot, dry and full of eucalyptuses. Looking forward to the challenge and change!

Wednesday, August 20, 2008

Going Down Under

Winter has arrived early for Miina and me this year. Although it has not directly been snowing on any member of the LAI detectives team recently, chilly winds and sudden cold showers are typical for the season. The winter did not come suddenly, though. The first public sign was a laconic post in the News section of the LAI detectives website in January: "Matti receives a Go8 Fellowship!". Miina joined in on the travel, sponsored by her generous Academy of Finland grant. As an consequence, tension has been building up since May as many small and large businesses needed to be taken care of before the departure: the PHYSENSE seminar, discussing with visiting Yuri Knyazikhin our latest ideas, moving between Helsinki and Tartu, making travel arrangements, finishing manuscripts and presentations, and, last but not least, being on a summer holiday.

Certainly we felt an amount of relief when the door of the airplane closed and we were on our way to Cairns, Queensland, Australia. Weatherwise, this was a nice continuation (or a culmination) to the unusually cool summer in Estonia. Although the nights were chilly in Cairns, there was still enough sunshine during the daytime to make up for it. Fortunately beach weather was kept outside the airconitioned premises of the conference center where we attended AGU Western Pacific Geophysics Meeting.

Alas, the tropical Far North was only a stopover on our journey further south. We said good-bye to the fine weather, sandy beaches, rainforests (covering the hills and emitting smoke as the backdrop to the kangaroos in the photo above) and the Great Barrier Reef, and boarded a plane again.

The final destination and our city of residence for the next four months was Adelaide, the capital of South Australia. At the airport we were warmly greeted by our host Prof. Megan Lewis from the University of Adelaide. Despite the friendliness of the people around here, it is difficult to get used to the abrupt ending of a cool northern summer, especially after the break in Queensland. The fact that this is an unusually cold winter in Australia and I have very little warm clothes in my 20 kg of luggage is not making things easier. Also, as the days were still long in Estonia, the rainy and cool days with sunset before 6 p.m. feels like the dark ages are back. However, these cold days are not to last and there is already a hint of spring in the air, in the songs of the birds and the smell of the flowers. No worries!

River Torrens just north of the center of Adelaide on a cold winter's day.

Sunday, July 20, 2008

Heroes of forest inventory

If you have ever headed back to forest after dinner feeling that your sample plot quota for the day has not been fulfilled, this post is devoted to you. You are one of the heroes of forest inventory.

One very typical feature in planning of different forest inventories is that the amount of work needed to finish the campaign is underestimated. Usually the result is that the number of plots or measurements that have been made in the given time is considerably smaller than was planned. However, sometimes it happens that the field crew won’t submit to this, but finds extra energy to carry on with the work when others would have given up. In Joensuu, by far the most legendary example of this are the 474 Matalansalo sample plots, measured in August/September 2004 by Aki and Henri. Their epic work has thereafter inspired many Joensuu’s forest inventory students, and the data has been utilized in numerous studies in the field of airborne laser scanning.

What makes some of us capable of heroic inventory achievements? In practice, forest inventory work means long days in hostile forests, with an immense herd of blood-thirsty mosquitoes whining in your ears, dry spruce twigs stabbing your face, and rain or sweat soaking you wet every day. My feeling is that the motives include at least an extra-developed sense of responsibility and a will to show your superiors that you can reach the goals that you were given. What is clear is that this kind of behaviour is most typical to students, especially if they are gathering data for their own research project. The students at this summer's MARV1-course in Hyytiälä showed that the tradition is likely to continue – hopefully their extra motivation did not arise from pure compulsion, but from knowing that their data will used in real research. On the other hand, more senior field workers seldom bother to continue after their working hours are done – they do the work they are paid for properly and deserve all the respect, but researchers that have the courage to employ students may experience also positive surprises…

And what about my mensuration campaign this summer? The fact that I needed to take two extra weeks in Hyytiälä to finish all the planned measurements should tell everything.

Never let someone else choose your plot locations!

Field assistant establishing LAI-plot

Monday, July 7, 2008

Breaking the routine

In April, Lauri described a typical week in office and lunch as the highlight of a day. I guess this is how many of our days are, nothing too much happening. Fortunately, there are also times when I wish I would have just one ordinary day in office to take care of this and that - the normal, boring days in office are after all something I enjoy.

I spent most of May and June in Helsinki, and it proved to be time for flying from one place to another - no time for sitting in the office and staring at my computer screen. There were two main reasons for this: the PHYSENSE workshop we were organizing in Helsinki and the Solemn Conferment ceremonies of our Faculty (and all the preparations related to both of them).

The PHYSENSE workshop was the second one we organized, so there was a certain routine to it already. Things went smoothly, and at least I enjoyed the workshop and meeting friends once again.

The Solemn Conferment ceremonies at our Faculty are arranged every six years and last for three days including two formal dinners, the Conferment, a grand ball and a countryside picnic. The dress code for all the events is very strict, and takes some time (and effort) to fulfill. Through May, we also practiced old academic dances such as Krakowiak, Valse Mignon, Cicapo and Pas d'Espagne. However, for me, the toughest part was preparing my two (trilingual) speeches for the Solemn Conferment. I know it is the daily job of many people to make formal speeches for big audiences, but for a regular postdoc like me it is certainly a task out of the ordinary. I spent two weeks before the ceremonies simply being nervous...

During the Solemn Conferment, a special reason for joy for our research group was that Tiit Nilson was appointed as an honorary doctor in our Faculty - a beautiful way to acknowledge his long collaboration with the Finnish forest research community.

Recovering from the Solemn Conferment (and the four hours of sleep per night we were able to catch) took me several days. I can honestly admit that I was sleeping in front of my screen, not just staring at it.

A few days after the celebrations Yuri Knyazikhin came for a visit. It was very nice seeing him again, and hearing about future research plans and planning collaboration. In our mini-seminar (i.e. Yuri, Pola, Matti, me) Yuri presented his ideas which involved (surprise, surprise!) our favorite spectral invariants, and Matti presented his new algorithm for retrieving leaf albedo from hyperspectral RS data.

After waving goodbye to Yuri, it was the time for me to say goodbye to Helsinki and head for Tartu (and summer vacation!). Ordinary office days are getting closer...

Conferment picnic (by Matti)

Wednesday, July 2, 2008

PHYSENSE workshop

We had the second PHYSENSE workshop in Helsinki in the beginning of June. Lots of interesting talks and new faces! I was personally very glad to see so many (young) people presenting their work. (For those interested, the abstracts of the PHYSENSE talks can be accessed at a website set up by Lauri.)

Saturday, April 19, 2008

A week in office

Based on this blog you might think that nothing really significant has happened in the first months of this year. Well, I can tell you the truth: nothing really significant has happened. People are mostly just sitting in their offices doing whatever small things are part of their research work. But maybe even the ordinary can be worth reporting - once.

Monday. Rain and snow. I spent the morning making figures for the next manuscript. In the afternoon I browsed through one biometrics book wondering if I need to include some kind of statistical test to confirm that my results are what they look like. Rest of the day was mainly putting the results to paper, one page more was finished. Highlight of the day: final program of the physense-seminar was announced.

Tuesday. The proofs of a recently accepted laser manuscript (codename: motolaser) arrived for checking and final corrections. Most of the day was spent in this job. I also managed to add a few more paragraphs to the yesterdays work before leaving to the sunny spring day outside. Highlight of the day: hopefully this really was the end of the motolaser project, something that started already in fall 2006 and has required a remarkable amount of time and patience since then – while being really funny at times.

Wednesday. Continuing with the manuscript. I decided that one of the figures needs more points, which meant a significant workload for my computer during the morning. In the afternoon I continued writing, and managed to finish early versions of both results and discussion sections. Highlight of the day: an interesting discussion about one particular pine tree growing more than 200 km away.

Thursday. Wrote abstract to the article and spent rest of the day reading once again those same old, vague articles to confirm that I had cited them correctly. Highlight of the day: lunch break.

Friday. Another sunny but quite boring day. More checking of references and clarifying the unclear parts of the manuscript. Happily all this was finished by the end of the day so the paper was more or less ready to move forward. Highlight of the day: closing the pc for weekend.

Friday, February 15, 2008

Cajanus tube demystified

One of the questions I keep on hearing is: what is the Cajanus tube and how it is used? Actually it is all very simple, but in scientific articles there is never enough space to describe everything comprehensively. So hopefully this will make things clearer.

At least in this case one image tells more than tens of words, so this is how the Cajanus tube looks like:

As you see, it is little like a periscope, but it is used for looking upwards. The balancing system at the top keeps the tube leveled, and the user looks vertically through the tube that has a small mirror (placed at 45 degrees angle) at the bottom and a crosshair at the top. The whole system is usually placed on a supportive staff. This allows the measurer to define exactly whether some point is covered by the canopy. It can be used for measuring crown radiuses or canopy cover. In canopy cover estimation the standard method is to establish some transects in the study plot, and then measure the percentage of the transect covered by the crowns.

The Cajanus tube has already a long history. It was designed by Werner Cajanus, who was the first professor in forest inventory in Finland, already in the 1910’s for use in crown radius measurements. Risto Sarvas showed how it can be used for canopy cover measurements in the 50’s, but after that I have not found any Finnish publications until LAI detectives. Actually nobody of our forest inventory staff here in Joensuu had even seen a Cajanus tube before I borrowed the one in the picture from Metla’s Suonenjoki research station.

The final question of an interested reader should now be “where to get one?” As far as I know, nobody makes these anymore. There is one fairly similar commercial version called the GRS densitometer, but it is nowhere as useful as the Cajanus tube, because it has no self-leveling system (the user must use the in-built leveling bubbles to keep it vertical). So the remaining option is to build one with your own hands – for example plywood and mirror can be used for the tube, and the leveling system can be made of some screws and iron bars. Or maybe an average researcher should just find someone handier to do the job…

Wednesday, January 2, 2008

America, America

December was an exciting month for me. Matti and I spent a week visiting Ranga Myneni's group at Boston University (BU) and another week in San Francisco participating in the fall meeting of the American Geophysical Union (AGU). I had been looking forward to the trip for quite a while - it would be nice to meet friends at BU (I was a visiting scholar there three years ago) and to go to a conference to get some new and fresh ideas. And of course, to see the way Americans prepare for Christmas (and all the other wintery holidays)...

Boston greeted us with wind and cold weather. Places looked pretty much the same as three years ago. I guess the biggest change was the new ticket system in the "T" (subway) and a Chinese extension to the campus food court. During our stay there, we had a very interesting discussion about p-theory and multi-angular data with Yuri and Mitch, and learned about the global greening and time trend research Sangram, Arindam and Ranga were working on. Sometimes even short, brief conversations can serve as a basis for starting something new. And hearing somebody else explain the thing you are working on, but with words different from those that you usually yourself use, can really help get a new perspective.

The AGU meeting in San Francisco was huge with about 15 000 participants and dozens of parallel sessions. For me, the most interesting part was a session dedicated to looking 50 years back and 50 years in to the future of Earth Observation missions. The session differed from the usual technical sessions; scientists who had worked some 30 years in the field talked about their experiences and thoughts for the future. My own contribution to the meeting was a poster presentation on estimation of photon recollision probability from multi-angular, hyperspectral data in the session "Remote Sensing of Land Surface Properties, Patterns, and Processes". The poster session was not as busy as I had hoped for, but to look for the positive side, nearly all the hand-outs I had prepared disappeared from my poster stand during the day.

Nights out in San Francisco were spent running around the city. Lots to see... the Sun setting in the Pacific Ocean, old cable cars, the Golden Gate Bridge, lazy California sea lions at Pier 39, the crooked Lombard street, busy Chinatown, Castro and Haight-Ashbury. Overall, the trip to "America" was a success, and my book shelves now have 10 kg's more books than before the trip! And for some mysterious reason, my jeans have shrunk.

San Francisco streets.