Select fields · 2 selected
Field settings 1 day · mean · obs · 0–0.2 fraction
Source & coverage
UCAR / CU · CYGNSSEstimated surface-water fraction from CYGNSS observations for one day. Transparent areas can mean missing data or values below the display threshold.
Field settings 1 day · mean · obs · 0–20 mm/day
Source & coverage
UCSB / CHC · CHIRPS v3 (IMERG daily)CHIRPS daily land precipitation. Displayed in mm/day; Sum gives accumulated rainfall over the selected period.
Prepared fields
Prepare the selected layers for the displayed date and preceding playback dates. Up to 14 days, 96 field requests and 64 MB; prepared responses expire after five minutes.
0/0 prepared · 0 unavailable · 0 cached · 0.0 MB
2 selected fields
Help
Select any number of fields, set their opacity and use the arrows to change their order. Higher fields draw above lower fields. In two-map view, assign each field to Map 1, Map 2 or Both. The background stays at the bottom.
The header date is the end of the trailing Window. Each field uses its selected Mean, Min, Max or Sum. Rainfall presets have independent 1-, 3- or 7-day windows. Field settings contains level, dates, anomalies, scale, colours/contours and transparency controls. Changes load automatically. Back/Forward restores previous setups.
Days before remains in Field settings → Date. Follow uses the header date; Fixed holds a field on its own date. Save JSON or Saved setups preserves the selected layers. Reset keeps saved cases; Clear fields and tracks keeps the date and map position.
Contours are sampled in a captured map region; use Use visible region after moving the map. Auto scale uses valid values in the selected field’s map. Export options applies to Save PNG and Print / PDF, including both maps. Preparation warms the selected layers for preceding playback dates with a bounded cache.
Hovmöller and Time series start with the first two selected changing numerical fields. Candidate markers and static references are not plottable. Missing observations remain blank. ERA5 spectral data cover 20°S–20°N; Wave Hovmöller is available from the header. Trajectories and seasonal ITCZ axes are selectable under Spectral fields; Track settings contains source, trail, threshold, extent and GeoJSON export.
DFO reported floods is available under CYGNSS. Its archive card browses 5,501 events from 1985–2023 and can load Spatial and Temporal comparisons for overlapping dates. DFO outlines, other flood-report polygons and Sumatra district-centre markers locate reported places, not measured flood footprints. CYGNSS Spatial and Temporal are algorithm candidates, not independently confirmed floods. Rivers and permanent water are static references; their source details are available in Field settings.
CYGNSS data
1-day: observed cells only. 3-day: rolling average with spatial interpolation. Transparent areas may be missing observations or inundation below 0.01. Dataset details ↗Save the current controls as a named setup. Load restores its controls and automatically loads its map. Setups are stored in this browser on this device.
Examples
Choose a map to load its region, date and selected layers. You can then adjust them. Time series opens prepared controls; click Plot to calculate.
DFO × CYGNSS · accumulated flood comparison
Compare Spatial or Temporal occupancy with DFO reported affected regions. Accumulate five years, the full shared archive, individual years or custom dates. Counts distinguish daily patches from linked episodes; click a DFO event to inspect daily CYGNSS observations.
Spatial candidate hotspots · 2020–2025
All qualifying daily patches, including patches without a temporal link. Explore the global and tropical top-eight rankings, then compare with Temporal using Candidate definition. Regional markers cover a selected 30-day inspection period; percentages summarize 2020–2025. Figure PDFs, notebook and result tables are linked in the region panel.
Temporal candidate hotspots · 2020–2025
Locate the 14 distinct 5° boxes in the global and tropical rankings. Percentages measure candidate cell-days among valid observations, not flood-disaster frequency. Regional views use the 30-day window with the most candidate cell-days; this selection helps inspection and is not independent validation.
Sumatra flood reports · 9 March 2023
Compare orange BNPB circles with cyan circles from the Sumatra newspaper archive. Circles show district centres; shared dates and districts do not prove that reports describe the same flood.
Okawango delta · Botswana · 2022
Compare two seasons with the same inundation colour scale, then explore rainfall farther north. Rectangular time-series means are separate from catchment delineation.
Eastern Australia · 2022
Compare CYGNSS inundation in the same region and on the same scale. The time series compares inundation and rainfall on separate axes.
North Sumatra · Indonesia · November 2025
Compare algorithm candidates with Groundsource news-location polygons. Two linked candidate anchors fall within reported locations on their observation dates; this does not prove identical flood footprints.
Dates and windows
Header date: UTC. Choose each field’s window and calculation together in Field settings.
Each field has its own trailing window, including its end date, and may use a fixed date or end earlier using Days before. The summary gives its actual date range. Rainfall totals are millimetres; daily rainfall is mm/day. Static reference layers do not change with the selected date.
The ERA5 interval chooses one six-hour interval for a one-day spectral field. Longer windows use the available intervals across the selected calendar days. Missing data are not filled with another date or pressure level.