Water Dynamics Analysis in Karst Flood Areas Using Sentinel-1 Time Series

Breznik, Jana (author), Oštir, Krištof (author), Ivačič, Matjaž (author), Rak, Gašper (author)

Studying karst water dynamics is challenging because of the often unknown underground flows. Therefore, studies of visible karst waters receive considerable research emphasis. Researchers are turning to various data sources, including remote sensing imagery, to study them. This research paper presents an assessment of a water bodies dataset, automatically detected from Sentinel-1 imagery, for karst flood research. Statistical and visual analyses were conducted to assess the reliability and effectiveness of the dataset. Spearman’s correlation coefficients were employed for statistical analysis to determine the degree of correlation between the areas of water bodies dataset and official water level data. Visual analyses involved the creation of heat maps based on the identified water areas, which were then compared to official flood maps, and the preparation of an analysis of historical flood events or results of hydrological and hydraulic modelling. Additionally, vegetation maps were produced to identify areas that lacked detection and complemented the heat maps. Statistical assessment showed a strong correlation (≥0.6) between the dataset and official water level data in smaller flood-prone areas with less complex inflow. Visual analyses using heat maps and vegetation maps effectively identified frequently flooded areas but had limitations in areas with dense vegetation. Comparisons with flood maps showed an important value of the dataset as an additional source of information for karst flood studies. This assessment highlights the dataset’s potential in combination with other data sources and modelling approaches.

• Language: English
• Keywords: Sentinel-1, radar satellite images, SAR, flood areas, water dynamics, karst
• Work type: Article
• Typology: 1.01 - Original Scientific Article
• Organization: FGG - Faculty of Civil and Geodetic Engineering
• Publication status: Published
• Publication version: Author Accepted Manuscript
• Year: 2023
• Number of pages: 26 str.
• Numbering: Vol. 15, iss. 15, art. 3861
• PID: 20.500.12556/RUL-151780
• UDC: 528.7:556:531/.532(045)
• ISSN on article: 2072-4292
• DOI: 10.3390/rs15153861
• COBISS.SI-ID: 161556995
• Publication date in RUL: 20.10.2023

Record is a part of a journal

• Title: Remote sensing
• Shortened title: Remote sens.
• Publisher: MDPI
• ISSN: 2072-4292
• COBISS.SI-ID: 32345133

Licences

• License: CC BY 4.0, Creative Commons Attribution 4.0 International
• Link: http://creativecommons.org/licenses/by/4.0/
• Description: This is the standard Creative Commons license that gives others maximum freedom to do what they want with the work as long as they credit the author.

Secondary language

• Language: Slovenian
• Title: Analiza dinamike vode na kraških poplavnih območjih z uporabo časovnih serij Sentinel-1

Abstract

Raziskovanje dinamike kraške vode je zaradi pogosto nepoznanih podzemnih tokov zahtevno. Posledično je raziskavam vidnih kraških voda namenjen velik raziskovalni poudarek. Raziskovalci se pri njihovem preučevanju obračajo k različnim virom podatkov, vključno s posnetki daljinskega zaznavanja. V tem raziskovalnem članku je predstavljena ocena nabora podatkov o vodnih telesih, samodejno zaznanih iz posnetkov Sentinel-1, za raziskave kraških poplav. Izvedene so bile statistične in vizualne analize za oceno zanesljivosti in učinkovitosti nabora podatkov. Za statistično analizo je bil uporabljen Spearmanov korelacijski koeficient, s katerim je bila določena stopnja korelacije med nizom podatkov o vodnih telesih in uradnimi podatki o vodostaju. Vizualne analize so vključevale izdelavo toplotnih kart, izdelanih na podlagi prepoznanih vodnih območij. Te karte so bile v nadaljevanju primerjane z uradnimi poplavnimi kartami, pripravljenimi na podlagi analize preteklih poplavnih dogodkov ali rezultatov hidrološkega in hidravličnega modeliranja. Kot dopolnitev toplotnih kart so bile izdelane tudi vegetacijske karte z namenom opredelitve slabše zaznanih območij. Statistična ocena je pokazala močno korelacijo (≥ 0,6) med naborom podatkov in uradnimi podatki o vodostaju na manjših poplavnih območjih z manj kompleksnimi pritoki. Vizualne analize na podlagi toplotnih kart in dopolnitvijo vegetacijskih kart so učinkovito opredelile pogosto poplavljena območja. Izkazala se je slabost zaznavanja na območjih z gosto vegetacijo. Primerjave s poplavnimi kartami so pokazale pomembno vrednost podatkovnega niza kot dodatnega vira informacij za kraške poplavne študije. Izvedena študija poudarja potencial podatkovnega niza v kombinaciji z drugimi viri podatkov.

• Keywords: Sentinel 1, radarski satelitski posnetki, SAR, poplavna območja, dinamika voda, kras

Projects

• Funder: ARRS - Slovenian Research Agency
• Project number: P2-0406
• Name: Opazovanje Zemlje in geoinformatika