Google Earth engine (GEE) - global flood database V1 (2000-2018)

The global flood database contains 2000-2018 What happened in 913 The scope and time distribution of this flood event . For more information , see also   Relevant journal articles .

The flood event is from Dartmouth flood Observatory Collected , Used to collect MODIS Images . Selected 913 Events are those that use Terra and Aqua MODIS Sensors 12,719 Events successfully drawn in scenarios （ Through quality control , Because there is obvious inundation beyond the permanent water ）. Within the whole date range of each flood event , With 250 The resolution of meters classifies each pixel as water or non-water , And generate subsequent data products , Including the maximum flood range （“ The flood ” belt ） And inundation duration in days （“ The duration of the “ Band ）. Water and non-water categories during flood events include permanent water （ Resample here 30 rice  JRC Global surface water data set ）  Represents permanent water , A resolution of 250 rice ）, have access to “jrc_perm_water” The wave band shields it to isolate floods . Added additional data quality bands , Indicates the cloud condition during a flood event （ for example ,“clear_views” Indicates the number of sunny days observed between the flood start and end dates ,“clear_perc” Indicates the percentage duration of sunny observation in the whole event （ God ）.

ImageCollection Each image in represents a map of a single flood . The set can be by date 、 National or Dartmouth flood observatory original ID To filter .

Dataset availability

2000-02-17T00:00:00Z - 2018-12-10T00:00:00

Data set provider

Cloud to Street (C2S) / Dartmouth Flood Observatory (DFO)

Earth engine references ：

ee.ImageCollection("GLOBAL_FLOOD_DB/MODIS_EVENTS/V1")

The resolution of the
30 rice

Band

full name	Company	Sensitive	maximum	describe
flooded		0	1	The maximum extent of flooding during the event . 1 - Surface water area 0 - There is no water
duration	God			Duration of surface water during the event （ In days ）. The pixel value represents the composite number of days that the pixel area is treated as water during the event . Use for 3 Days of MODIS Composite image .
clear_views	God	0	65535	The number of cloudless observation days between the start and end dates of each event . Cloud coverage is determined by MODIS Quality assurance belt （'state_1km'） determine .
clear_perc	%	0	100	Percentage of clear field observations during a given flood event . This is equivalent to “clear_views” Band , But normalized to each flood event MODIS Number of images . Cloud coverage is determined by MODIS Quality assurance belt （'state_1km'） determine .
jrc_perm_water		0	1	from JRC The global surface water dataset uses “ transition ” Permanent water with defined wave band . Keep the resolution at JRC Raw data set 30 Meter resolution . 1 - Permanent water 0 - Non water .

Image properties

full name	type	describe
ID	INT	With Dartmouth flood observatory (DFO) A unique directory of consistent flood events ID.
CC	String	With the DFO Event polygons intersect the watersheds in which floods are detected / Three letters of the region ISO Country / Region code （ In the list ）.
Country	String	With the DFO The country name of the country where the flood was detected in the watershed intersected by the event polygon （ In the list ）.
dfo_centroid_x	Double that	To estimate the location of a flood event DFO The longitude of the centroid of a polygon （DFO database ）.
dfo_centroid_y	Double that	To estimate the location of a flood event DFO The centroid latitude of a polygon （DFO database ）.
dfo_country	String	Major flood countries （DFO database ）.
dfo_other_country	String	The secondary country of the flood （DFO database ）.
dfo_displaced	INT	The estimated total number of people displaced or evacuated after the flood event （DFO database ）.
dfo_main_cause	String	DFO The main causes of flood events in the database . Not standardized .
dfo_severity	Double that	The severity of the flood event （DFO database ）： 1 - Large flood events 、 Major damage to structure or agriculture 、 Death toll and / Or reported since the last similar incident 5-15 Annual interval 1.5 - A very big event ：>15 Nian Dan <100 Recurrence interval of years 2 - Extreme events ： Recurrence interval > 100 year ）
dfo_dead	INT	The estimated number of deaths caused by the flood event （DFO database ）.
dfo_validation_type	String	The main source of flood event confirmation （DFO database ）. Not standardized .
glide_index	String	Global identifier number .
gfd_country_code	String	Countries that intersect with watersheds / Comma separated two letters of region FIPS Country / List of area codes , The watershed is used as the region of interest in the water detection algorithm .
gfd_country_name	String	The country name of the country that intersects the watershed used as the region of interest in the water detection algorithm （ In the list ）.
The compound type	String	The number of days used for synthesis in the water detection algorithm .
Threshold type	String	Used to classify water / Threshold type of non-aqueous detection algorithm - “otsu” or “ standard ”.
threshold _b1b2	Double that	Used in water detection algorithm b2b1 Threshold of the ratio .
threshold _b7	Double that	Applied to the band used in the water detection algorithm 7 (SWIR) The threshold of .
otsu_sample_res	Double that	Used to build MODIS The spatial resolution of the mosaic reducer （ With m In units of ）, Then sample and estimate otsu threshold （ For use only otsu Instead of the default threshold flood event ）.
Slope threshold	Double that	The value used to mask the steep region from the water detection algorithm , To minimize the error of terrain shadow .

Code ：

var gfd = ee.ImageCollection('GLOBAL_FLOOD_DB/MODIS_EVENTS/V1');

//  A single flood event -- Floods caused by Hurricane Isaac in the United States .
var hurricaneIsaacDartmouthId = 3977;
var hurricaneIsaacUsa = ee.Image(
    gfd.filterMetadata('id', 'equals', hurricaneIsaacDartmouthId).first());
Map.setOptions('SATELLITE');
Map.setCenter(-90.2922, 29.4064, 9);
Map.addLayer(
  hurricaneIsaacUsa.select('flooded').selfMask(),
  {min: 0, max: 1, palette: '001133'},
  'Hurricane Isaac - Inundation Extent');

//  The duration of the （ The number of days the flood event lasts ）.
var durationPalette = ['C3EFFE', '1341E8', '051CB0', '001133'];
Map.addLayer(
  hurricaneIsaacUsa.select('duration').selfMask(),
  {min: 0, max: 4, palette: durationPalette},
  'Hurricane Isaac - Duration');

//  Map all floods , To generate historical flooding areas observed by satellites .
var gfdFloodedSum = gfd.select('flooded').sum();
Map.addLayer(
  gfdFloodedSum.selfMask(),
  {min: 0, max: 10, palette: durationPalette},
  'GFD Satellite Observed Flood Plain');

//  Cover permanent waters to distinguish floods .
var jrc = gfd.select('jrc_perm_water').sum().gte(1);
Map.addLayer(
  jrc.selfMask(),
  {min: 0, max: 1, palette: 'C3EFFE'},
  'JRC Permanent Water');

  quote ：

Citations:

• Tellman, B., J.A. Sullivan, C. Kuhn, A.J. Kettner, C.S. Doyle, G.R. Brakenridge, T. Erickson, D.A. Slayback. (Accepted.) Satellites observe increasing proportion of population exposed to floods. Nature. doi:10.1038/s41586-021-03695-w

result ：