Spaces:
Running
Running
| # """Documentation page for Streamlit app.""" | |
| import streamlit as st | |
| from src.config_parameters import params | |
| from src.utils import ( | |
| add_about, | |
| set_doc_page_style, | |
| toggle_menu_button, | |
| ) | |
| # Page configuration | |
| st.set_page_config(layout="wide", page_title=params["browser_title"]) | |
| # If app is deployed hide menu button | |
| toggle_menu_button() | |
| # Create sidebar | |
| add_about() | |
| # Set page style | |
| set_doc_page_style() | |
| # Page title | |
| st.markdown("# Documentation") | |
| # First section | |
| st.markdown("## Methodology") | |
| st.markdown( | |
| """ | |
| This tool does not produce its own forecasts; it leverages the flood forecasts created | |
| by the GloFAS Global Flood Monitoring tool and aims to make them conveniently accessible. | |
| The GFM products are generated using flood detection algorithms applied to Sentinel-1 satellite data, | |
| which captures radar imagery in all weather conditions. Sentinel-1 data, | |
| acquired in Interferometric Wide-swath mode and VV-polarization, is preprocessed into Analysis-Ready Data | |
| (ARD) with a 10x10 m pixel resolution. Three flood detection algorithms are then run in parallel on | |
| this ARD: | |
| - **HASARD (by LIST)**: Uses image comparison and statistical modeling to detect changes in | |
| flood-related signals. | |
| - **Alg2 (by DLR)**: Applies fuzzy logic and hierarchical thresholding to classify flooded areas. | |
| - **Alg3 (by TUW)**: Leverages long-term signal history and statistical modeling for efficient | |
| global flood mapping. | |
| Each algorithm independently classifies flooded pixels, and their results are | |
| combined into a consensus map. A pixel is marked as flooded if at least two of the three algorithms | |
| agree. This ensemble approach improves accuracy and ensures near-real-time flood monitoring globally. | |
| Detailed documentation on the methodology is available on the GloFAS | |
| website: https://global-flood.emergency.copernicus.eu/technical-information/glofas-gfm/ | |
| The GloFAS documentation mentions 11 products that are published. The products used in this tool are | |
| - **The observed flood extent**: these are the floods shown in red when analyzing floods on the | |
| "Flood Analysis" page | |
| - **The Sentinel-1 footprint**: this is the bounding box of the Sentinel-1 satellite image that contains the | |
| flood, shown in yellow when analyzing floods | |
| """ | |
| ) | |
| # Second section | |
| st.markdown("## Radar imagery for flood detection") | |
| st.markdown( | |
| """ | |
| As described above, GFM uses Sentinel-1 data as the basis for its flood forecast. | |
| Sentinel-1 data is the result of measurements from a constellation of two | |
| satellites, assing over the same areas following the same orbit on average | |
| every 6 days. The figure below shows an overview of the Sentinel-1 observation | |
| plan, where pass directions and coverage frequencies are highlighted. | |
| More detailed documentation on Sentinel-1 can be found on the Copernicus website: | |
| https://sentiwiki.copernicus.eu/web/sentinel-1 | |
| """, | |
| unsafe_allow_html=True, | |
| ) | |
| # Add image satellite overview | |
| st.image( | |
| "%s" % params["url_sentinel_img"], | |
| width=1000, | |
| ) | |