User Interface
==============

The FloodCast web dashboard is accessible at :extlink:`https://floodcast.hpc.cineca.it/`.
No authentication is currently required.
It provides two main views corresponding to the two product types.

.. contents:: In this section
   :depth: 2
   :local:


Streamflow View
---------------

The streamflow view displays all FloodPROOFS river sections on an interactive map.

.. image:: _static/images/streamflow.png
   :alt: Streamflow map view showing river sections across Italy

Marker Appearance
~~~~~~~~~~~~~~~~~

Each section is represented by a marker whose shape and colour indicate the current
risk level:

* **Low Risk** — flow within normal range
* **Watch Level** — elevated flow conditions (threshold Level 1 exceeded)
* **Moderate Risk** — high flow with flood potential (threshold Level 2 exceeded)
* **Severe Risk** — critical flood conditions (threshold Level 3 exceeded)

Alert thresholds (Level 1, Level 2, Level 3) are collected from local authorities and
are defined on a per-station basis. Not all stations are provided with threshold values.

Map Interaction
~~~~~~~~~~~~~~~

Users can:

* Zoom in/out to explore specific areas
* Pan across the map
* Click markers to select monitoring stations

Control Panels
~~~~~~~~~~~~~~

The left-hand control panels provide:

* **Model selection** — choose between IFS and ICON (when available for the selected date)
* **Map style** — Carto Light, Satellite Hybrid, Dark Terrain
* **Date selector** — calendar widget where bold dates indicate FloodPROOFS output is
  available, greyed-out dates indicate no data, and the most recent forecast is shown
  by default
* **Download** — button to download the NetCDF streamflow file for the selected model and date
* **Legend** — marker shape and colour indicate risk level

.. image:: _static/images/Dates_available.png
   :alt: Date selector calendar with bold dates indicating available forecasts


Time Series Plot
~~~~~~~~~~~~~~~~

Clicking a station marker opens a detail panel displaying:

* **Chart**: simulated discharge (m³/s) over the forecast period (hours),
  with alert-threshold levels (Level 1, 2, 3) as horizontal dashed lines where available
* **Header**: station name, risk level badge, geographic coordinates, municipality,
  province, region
* **Max predicted flow**: the peak discharge value over the forecast window
* **Model/date controls**: switch model or date within the panel
* **Key indicators**:

  * Correlation time (hours)
  * Calibration status (yes/no)
  * Forecast status (Low Risk, Watch, Moderate Risk, Severe Risk)

* **Download options**: CSV, JSON, or original NetCDF

.. image:: _static/images/station_detail.png
   :alt: Station detail panel showing discharge chart with threshold levels and download options


Download Options
~~~~~~~~~~~~~~~~

From the dashboard, users can download:

**Streamflow data:**

* **NetCDF file** — full forecast output for all ~500 stations for the selected
  model and date (via the download icon in the left panel)
* **CSV** — time series for a single station (from the station detail panel)
* **JSON** — time series for a single station (from the station detail panel)

**Return Period data:**

* **GeoTIFF** — the original single-band raster for the selected model and date,
  suitable for import into GIS software


Return Period View
------------------

The return-period view displays the maximum flood return period within the forecast
window for each pixel of the Italian river network.

.. image:: _static/images/return_period.png
   :alt: Return period map showing flood risk across Italian river network

Map Display
~~~~~~~~~~~

The map is colour-coded with return-period classes:

* T ≥ 2 years (green)
* T ≥ 5 years
* T ≥ 10 years
* T ≥ 20 years
* T ≥ 50 years (dark purple)

Control Panels
~~~~~~~~~~~~~~

* **Model selection** — IFS or ICON
* **Date selector** — same calendar behaviour as the streamflow view
* **Map style** — Carto Light, Satellite Hybrid, Dark Terrain
* **Download** — original single-band GeoTIFF file
* **Legend** — colour scale for return-period classes