NSK
NSK

Digitalni akademski repozitorij (DAR)

time: 0.0020320415496826
 
Naslov:Prepoznavanje konvektivnih oblaka, magle i niskih oblaka te lociranje šumskih požara multispektralnom analizom satelitskih slika : doktorski rad / Nataša Strelec Mahović
Ostali naslovi:Detecting convective clouds, fog and low clouds and locating forest fires using multispectral satellite image analysis
Vrsta:Disertacija, tekst
Područje: Geofizika
Datum objave: 2011
Jezik:hrvatski (hrv)
Sažetak (hr):
Tema ovog rada je poboljšanje, operativna primjena, testiranje i validacija metoda za prepoznavanje i praćenje razvoja konvektivnih oblaka, magle i niskih oblaka te lociranje šumskih požara, temeljenih na kombinacijama Meteosat satelitskih podataka u različitim spektralnim područjima. Istraživanje je podijeljeno u tri zasebne tematske cjeline, od kojih je najdetaljnije obrađena tema praćenja razvoja konvektivnih oblaka. Poglavlje 2 donosi osvrt na dosad korištene metode prepoznavanja konvekcije i njihove nedostatke. Zbog potrebe za poboljšanjem izveden je novi produkt, temeljen na razlici refleksivnosti u kanalima 0.6 i 3.9 μm. Pokazuje se da ta razlika prepoznaje male konvektivne stanice u početnoj fazi razvoja. Promjena razlike refleksivnosti, s početnih 40 ili 50% na 60 ili 70% nakon 15 minuta, najčešće se može povezati s naglim razvojem konvektivne stanice. Razlika refleksivnosti veća od 80% javlja se kad su na vrhu oblaka prisutni veliki kristali leda, što je pokazatelj zrele faze konvektivne stanice. Također su istražene karakteristike vrhova konvektivnih oblaka uočene u različitim spektralnim područjima. Primjećeno je da se područje najveće razlike refleksivnosti u kanalima 0.6 i 3.9 μm, kao i područje najvećeg albeda u kanalu 0.6 μm, nalaze jugozapadno od najhladnijeg dijela oblaka. Istodobno se najhladniji dio oblaka poklapa s područjem najveće refleksivnosti u kanalu 3.9 μm. Povećana refleksivnost u kanalu 3.9 μm povezana je s jakom uzlaznom strujom koja na vrh oblaka izbacuje male kristaliće leda. Iz svih obrađenih primjera može se zaključiti da je najaktivniji dio oblaka onaj u kojem je razlika refleksivnosti u kanalima 0.6 i 3.9 μm između 60 i 80%. Za prepoznavanje magle i niskih oblaka, obrađeno u poglavlju 3, noću se koristi razlika temperature u kanalima 10.8 i 3.9 μm, uz graničnu vrijednost razlike 3 K, dok je danju uključen i kanal 1.6 μm. Usporedba s motrenjima na postajama pokazuje uspješnost metoda. S obzirom na to da satelit ne razlikuje maglu od niskih oblaka uveden je dodatni kriterij u kojem se temperatura u kanalu 10.8 μm uspoređuje s mjerenom temperaturom na 2 m. U lociranju šumskih požara kljucnu ulogu ima temperatura u kanalu 3.9 μm, kao što je pokazano u poglavlju 4. Kao dodatni kriterij uzima se razlika temperature u kanalima 3.9 i 10.8 μm. Obrađeni podaci o svim požarima većim od 1 ha, tijekom požarne sezone 2009., pokazuju da je samo 7% požara bilo uočljivo u satelitskim podacima. Ostali nisu bili vidljivi uglavnom zbog postojanja oblaka iznad požarišta ili premale površine požarišta. Ovakav rezultat ukazuje na to da sustav za rano upozoravanje na požare ne može biti temeljen samo na satelitskim podacima s geostacionarnog satelita zbog premale vjerojatnosti detekcije, osobito kad se radi o malim požarima. Na kraju se može zaključiti da je od tri prikazane mogućnosti korištenja kombinirane analize satelitskih slika najuspješnija ona u prepoznavanju i praćenju konvektivnih oblaka.
Sažetak (en):
The topic of this dissertation is improvement, operative implementation, testing and validation of methods for detection and monitoring of convective clouds, fog and low clouds and locating forest fires, using a combination of Meteosat satellite data in different spectral channels. The study was divided into three separate thematic units, of which the most detailed, presented in Chapter 2, deals with monitoring of convective clouds. The goal of the research was to provide operational forecasters with satellite products that would enable easier recognition and nowcast of typical important weather phenomena. Detection of potentially dangerous convective clouds is one of the most important tasks in operational weather services. Low predictability of the processes causing convective development emphasizes the importance of the methods for early detection and nowcast of convective cells. Satellite-based methods play a major role in this field. Chapter 2 explains the possibilities of using satellite data in the analysis of all stages of convective development. In pre-convective stage, temperature difference of channels 6.2 and 7.3 μm locates the unstable regions. Concerning the early detection of convective cells, this work makes reference to the previously used methods of convective clouds recognition, based on 10.8 μm channel data, and discusses their shortcomings. Following the need to improve the existing methods, a new product is designed, based on reflectivity difference of 0.6 and 3.9 μm channels. Validation of the method includes comparison to radar and lightning data. The results show that the reflectivity difference detects small convective cells in the initial stage of their development. Changes in difference levels, from the initial 40% or 50% to 60 or 70% after 15 minutes, can be associated with the rapid development of a convective cell. The difference values of 80% or larger are connected with large ice particles on top of convective cloud, absorbing most of the solar radiation in 3.9 μm channel. These are the values that can be associated with mature convective cells. Additionally, analysis of cloud-top characteristics in channels 0.6, 3.9 and 10.8 μm as well as in 0.6 - 3.9 μm reflectivity difference was performed. It was noted that the area of highest 0.6 - 3.9 μm reflectivity difference, as well as the maximum of 0.6 μm reflectivity, are located southwest of the coldest part of the cloud. However, the coldest part of the cloud coincides with the area of highest 3.9 μm reflectivity. Increased 3.9 μm reflectivity can be associated with strong updrafts, bringing small ice particles to the top of the cloud. Finally, it can be concluded that the most active part of the cloud is the one with reflectivity difference of channels 0.6 and 3.9 μm between 60 and 80%. Chapter 3 is dedicated to the recognition of fog and low clouds. Method used for fog/low cloud detection at night utilizes the temperature difference of the channels 10.8 and 3.9 μm. During daytime, 1.6 μm channel reflectivity is added to the algorithm. Comparison with synop observations shows satisfactory performance of both methods, with limitations at the time of dawn and dusk. Given that the satellite cannot distinguish the fog from low clouds, an additional criterion has been introduced, comparing the 10.8 μm temperature with the measured 2 m temperatures. Chapter 4 presents the possibility of using satellite data to locate forest fires. The key in hot-spot detection is the 3.9 μm temperature, whereas an additional criterion is provided by the temperature difference of channels 3.9 and 10.8 μm. The analysis of all recorded fires larger than 1 ha, during fire season 2009, was performed. The results show that only 7% of fires were evident in satellite data. For all other fires detection was not possible. The most frequent reason were clouds above the fires, but in many cases fires were too small to be recognized and a large number of fires during dawn, dusk or even night was also not visible. This result indicates that the early warning system could not be based only on satellite data from geostationary satellites, since the probability of fire detection is too low, especially for small fires.
Brzi preglednik: ?vdoc=8809&page=1
big_sql: 0.0004119873046875 | sql_other: 0.0036499500274658