mfontenele

CBERS for Africa and CBERS-2B status

2009-05-23T00:38:00.004-03:00

From May 12 to 15, I've been to Cachoeira Paulista, SP to install 2 computer racks to be used in the CBERS for Africa project.

Getting the racks in the appointed place was not an easy task.

Each rack contains a complete system for CBERS reception. There are 3 PowerEdge servers, one for each channel: CCD1, CCD2 and HWDT-Q, a disk array and a tape changer, plus a desktop computer used for displaying the Moving Windows and for checking the satellite schedule.

We receive HWDT-Q in order to extract the WFI stream and GPS and Star Sensor data. The HRC stream won't be processed. In time the racks will be sent to Africa.

In his last trip to China, Lula signed an agreement for CBERS-2B reception in South Africa, the Canary Islands and Egypt. The Chinese are already receiving data in South Africa and our tests in Maspalomas (Canary Islands) and Aswan (Egypt) went well. There are still some tests to be made with the antenna in Maspalomas, they are scheduled for May 11-24. Unfortunately CBERS-2B entered safe mode on May 12, it appears its due to some problem in the attitude control system. Let's hope its fixable.

CBERS and me

2009-05-10T23:00:00.001-03:00

I think it's time for me to talk a little about how I began working with remote sensing.

It was the end of 1998 and I was just about to leave on vacation, after the end of the semester, when I met a friend that had just seen an employment announcement in our university job board. At the time, I had just left my former job to focus on finishing my degree and was not looking for a job at all.

According to him, a company was searching for people interested in working for the brazilian space program. In my mind the only space program we had was the military attempts to build a satellite launcher (a.k.a. ballistic missile). It never crossed my mind that we were almost ready to operate a remote sensing satellite. I didn't give it much thought and went to Cabo Frio for a season on the beach.

One weekend he came over and we talked for a long time about his new job and all the things he was learning. He said the company was still looking for programmers and asked if I was interested. I cut my vacation short and came back to Rio for the job interview.

The company was in reality a small room in an apartment building in front of the beach near my own home in Barra da Tijuca, a neighborhood in the west side of Rio. They had been doing system specification for a while and now we had to built it in 6 months. I would be working in Red-Hat Linux 4.2, a short time later migrating to 5, and programming in C. The system would later be deployed to a DEC Workstation. My previous experience was with C++ and Delphi on Windows and some C on AIX, but not much. I also had a little experience with the now infamous SCO Unix.

We were under contract to a french company named Matra Systems et Information (that later became Aérospatiale and finally EADS), developing the image processing sub-system for the CBERS, China-Brazil Earth Resources Satellite, ground station.

I began working for DVR (later Gisplan and finally AMS Kepler) in Jan 1999, CBERS-1 was successfully launched from China on October 14 of the same year. I was so nervous I can't even remember how the first reception went.

By the time CBERS-2, mostly a copy of CBERS-1, was getting ready for launch there was a big change in INPE's direction, with a focus on releasing satellite imagery for free using the internet. They also wanted the new system to be able to process there historical archives for MSS, TM and later to process LANDSAT-7 ETM+ data. We won the bid for the system, this time as the prime contractor, and the Multi-Satellite Station System, MS3, was born, but this is a topic for another post.

XIV Brazilian Remote Sensing Symposium

2009-05-04T23:26:00.010-03:00

I'm back from Natal, RN, Brazil, where I went for the XIV SBSR. This year's symposium was not as good as previous versions, maybe because the global economic crisis has reduced the number of sponsors.

The schedule was different from the other editions, in the mornings there was no break during sessions, it went straight from 9:00 to 12:00. In the past, there was a coffee break in the middle which allowed us to take a look at the exposition and to grab something to eat and go to the bathroom. Without the coffee break, the exposition was mostly deserted in the morning. This also happened because the exposition was at the back of the building, people could enter the convention center, go to their sessions, leave and never look at the exposition. If I remember correctly, you had to walk through the exposition in the last edition.

Another problem I had was that some of the sessions I wanted to watch happened at the same time, so I had to choose one and miss the other. It's hard to please everyone when doing a schedule and this time I was the unlucky part.

On a sort of positive note, for the first time, I heard an explanation for the problem with CBERS's CCD relative calibration. Flávio Ponzoni from INPE said, answering a question from the audience, that it seems as though the CCD camera is not radiometrically stable, that the detectors response vary from pass to pass making systematic relative calibration very difficult. Let's hope that the new sensors in CBERS-3 and CBERS-4 will do better.

On friday, May 1st, I was finally able to do some site seeing before catching my flight home. Together with some friends from USGS and NASA, and despite some rain in the morning, we went on a crazy buggy ride on the dunes. I highly recommend it, truly unforgettable.

Web Nameplate

2009-03-03T08:56:00.007-03:00

After reading the post Have a say in what Google says about you at Lifehacker, I decided to take matters into my own hands. My new personal web nameplate is at Marcelo Fontenele S Santos. There you can find some detailed info about me and contact me if you wish. At this time I am using Plaxo's public profile as the gateway.

Rotten Roots

2008-12-06T01:29:00.005-02:00

I have to agree with Jeff Atwood when he says that, in his opinion, software development is like tending a garden.

It's spring here in Rio and in every spring, for the last 3 years, I have changed the soil of my bonsai and increased the amount of water I give them. But this year the weather has been quiet different, for the most part it's been raining a lot, with a vicious wind, very unlike Rio.

Without enough time to tender my plants, the excess of water took its tool while I lay unaware. After two months of overwatering, I finally realized that two of my plants, the ones that I had put in a more organic soil mix, had their leaves turning yellow and soft - this was a bad sign. In a desperate act, I tried to change the soil to a pure inorganic compost but, as I removed the old soil, I realized that the roots of both plants were already completely rotten.

Let's get back to software, Andy Hunt and Dave Thomas in their book The Pragmatic Programmer talk about Broken Windows and Martin Fowler in Refactoring tells us to be wary of Bad Smells. The yellow and soft leaves could be seen as Bad Smells, as Broken Windows. When I saw those Broken Windows I tried to fix them. But unfortunately I was too late.

It's the same in software development, you have to be looking for those Bad Smells all the time. You must fix the Broken Windows as soon as they appear or it could be too late.

As with my plants, there are always external factors that must be taken into account when developing a system. It could be a changed requirement or a lack of communication between the team or with a client or a thousand other things. The environment surrounding the software have a direct influence in the outcome of a project and it must be taken into account, every change must be studied so we can measure its impact.

A strong wind may brake a lot of windows.

GEO announces free Landsat imagery

2008-11-26T22:26:00.021-02:00

The Group on Earth Observations (GEO) announced on 20 November 2008 that "the world will soon have the unrestricted access at no charge to the Landsat archive" .

Remote-sensing satellites are impartial and essential recorders of the fast-moving story of the Earth’s changing surface,” said José Achache, Director of the GEO Secretariat. “Landsat’s nearly four decades of accumulated Earth imagery data will provide an historical record that, combined with continuous updates, will make it possible to interpret and anticipate changes to the Earth’s surface with far greater certainty than ever before.

It' s good to see that they recognized the pioneering role of INPE.

GEO’s announcement last year that the China-Brazil Earth Resources Satellites (CBERS) would distribute its images free-of-charge was an essential first step.

INPE's catalog has been distributing free imagery from the CBERS family of satellites since 2004 and has already made available the Brazilian historical archive of MSS, TM and ETM+ data.

During the last year we've been working with the USGS Eros Data Center to ensure that the geometric and radiometric quality of those images match their own products' and creating tools to enable the exchange of those scenes using a common data format.

The validation of this work is being done this week in Sioux Falls but the latest version of our ground-station software already delivers some of those improvements to all users of Landsat imagery.