Operational monitoring of snow cover using digital imagery
Loading...
URL
Journal Title
Journal ISSN
Volume Title
Sähkötekniikan korkeakoulu |
Master's thesis
Unless otherwise stated, all rights belong to the author. You may download, display and print this publication for Your own personal use. Commercial use is prohibited.
Authors
Date
2020-01-20
Department
Major/Subject
Space science and technology
Mcode
ELEC3039
Degree programme
Master’s Programme in Electronics and Nanotechnology (TS2013)
Language
en
Pages
70+22
Series
Abstract
Fractional snow cover (FSC) and snow depth (SD) are two important parameters used to calculate snow water equivalent and surface albedo, which are important physical quantities for applications in climatology, hydrology and meteorology. FSC is traditionally monitored using satellite data, but it is challenging for optical sensors to retrieve signals from the ground when forest canopy is present. Similar challenge exists for retrieving microwave signals from terrain with high slope rates. In addition to retrieval challenges, validation of FSC products are done using proxy parameters since in-situ FSC observations are very limited. This is because there are no devices or systems usable for continuous measurement of FSC and manual observation takes a lot effort and depends on subjective judgement. SD is traditionally observed by manual readings of snow sticks. Manual observations requires effort and presence of manpower, especially in remote areas. Also, temporal resolution of such observations are generally one day. In the last decades, manual observations are replaced with automated observations by ultrasonic and optical sensors in some countries, but the manual observation is still the primary method in many countries. Using webcam photography for environmental monitoring is an emerging method. During the latest years, numerous environmental camera networks are established in different parts of the globe. These networks offer high resolution digital imagery in high temporal resolution. More digital imagery is also available from cameras and camera networks established for other purposes, such as monitoring ski tracks, traffic, harbours, urban areas etc. It is previously studied that environmental parameters are observed from digital images using image processing methods. A novel system is previously introduced by Tanis et al. for automated monitoring of different parameters from multiple camera networks. This system allows acquisition of images from different sources by defining camera networks on a toolbox, so that it can process and visualise the images on a processing chain customised by input from the user via graphical user interface. The toolbox is called Finnish Meteorological Institute Image Processing Toolbox (FMIPROT). It can work also on cloud, to create automated and continuous processing of digital imagery. In this thesis, FSC and SD are estimated for multiple locations in Finland by processing images from MONIMET camera network for 2018 - 2019 season. Images are classified as snow covered or snow free in pixel level using an adaptive thresholding algorithm which determines a threshold value for the digital numbers (DN) of pixels in blue channel using histograms of the images. FSC is estimated by using snow presence in the pixels from the classification and spatial resolution of the pixels calculated from georectification of the images. Images are georectified using perspective projection. SD is estimated using an algorithm to find the intersection of snow surface and snow sticks by thresholding and segmentation. Estimations are assessed using observations from in-situ measurements and observations by visual inspection. FMIPROT processing system is deployed on cloud and the near real time (NRT) monitoring is set up for the same parameters in same locations. The processing is integrated into "FMIPROT & Camera Network Portal" website so that the visualised NRT results are available for public.Description
Supervisor
Rautiainen, MiinaThesis advisor
Arslan, Ali NadirKeywords
remote sensing, digital imagery, image processing, snow cover, cloud processing, environmental monitoring