Technical Notes

In this page, description of datasets and source codes are written.

Input Data

Description

FLOW method requires two input data. One is a fine-resolution flow direction map, and the other is a fine-resolution DEM. In tar archive, the flow direction map of GDBD (Global Drainage Basin Dataset) [Masutomi et al, 2009] and SRTM30 DEM [NASA] are included. SRTM30 DEM is modified for coordinating with the flow direction map of GDBD. Method for modification is described in the reference paper [Yamazaki, 2010]


FD-DEM_IMG
GDBD Flow Direction Map (left) and STRM30 DEM (right)

Those datasets are stored in FLOW/data/GDBD/ directory. Datasets are divided by each continent.

Continents and their ID number
ContinentAsiaAustraliaNorth AmericaSouth AmericaEuropeAfricaJapanKorea
Initialasaunasaeuafjpkr
ID number12345678

List of input datasets and their suffixes
SuffixDataFormat
.fdFlow Direction4-byte real
.elevmodElevation4-byte real
.lonlatLongitude/Latitude4-byte real
.upareaUpstream Drainage Area4-byte real
.GRDCsetPosition of GRDC StationText


All files except GRDCset is plain-binary file. In .fd file, flow direction is indicated by numbers as below:

Flow Directions
812
7 3
654

For example is the flow direction of a grid is indicated by 2, it is the flow toward northeast.


In .lonlat file, longitude is stored in 1st record, while latitude in 2nd record.
Columns in .GRDCset indicates GRDC-ID, X coordinate, Y coordinate, Longitude, Latitude, Upstream Drainage Area.


Please find detailed information in GDBD webpage.

Notice: Endian

All data is prepared in little-endian format. If your system is using big-endian format, you should convert the endian of input data by following steps:

% cd $FLOW/data/GDBD/
% make conv_endian
% chmod 755 endian.sh
% ./endian.sh

'$FLOW' is your FLOW directory.

Reference

If you use those input datasets for your research, please do not forget to include reference to the dataset: GDBD [Masutomi et al, 2010] and SRTM [NASA].


Development of highly accurate global polygonal drainage basin data

- Masutomi, Y., Y. Inui, K., Takahashi, and U. Matsuoka

Hydrological Procceses, vol.23, pp572-584, 2009

Output Data

Data Format

A river network map and its sub-grid topographic parameters are stored in plain-binary format. In case of default 1deg setting, there are 360 grids from west toward east direction starting from (179.5W), and 180 grids from north toward south starting from (89.5N).

Output Maps for use in CaMa-Flood
NameDataFormat
nextxy.binRiver Network mapInteger
elevtn.binChannel Elevation4-byte real
rivlen.binChannel Length4-byte real
nxtdxt.binDistance to downstream grid4-byte real
grarea.binUnit-catchment Area4-byte real
fldhgt.binFloodplain Elevation Profile4-byte real
GRDCxy.txtGRDC stations upstream gridsText


Downstream grid coordinate is stored in the river network map. X coordinate is stored in 1st record, and Y coordinate is stored in 2nd record.

FD-DEM_IMG
River network map derived by FLOW (blue).

make_lsmask.F

Land-Sea Mask

In make_lsmask.F, land sea mask is decided. If a gird has a GDBD pixel which has an upstream drainage area larger or equal to 20 km2 (you can change this number by editing 'minarea'), the grid is recognized as land. Continent ID numbers are allocated onto land grids, while 0 is given to sea grids.

const_network.F

In const_network.F, a river network map, channel altitude, channel length, and distance to downstream grid is decided.


Outlet Pixels and Channel Elevation

First, the outlet pixel of each grid-box is decided. The GDBD pixels with flow direction toward neighboring grid-box are chosen as potential outlet pixels. The potential outlet pixel with maximum upstream drainage area is selected for each grid, and the distance toward the selected potential outlet pixel of downstream grid is calculated. If the distance is smaller than the threshold value (default is 50 km, you can change the threshold by editing '_dst_min_' in 'define.h'), the selected potential outlet pixel is rejected. In case there is no other potential outlet pixel in the grid-box, then the selected outlet pixel in downstream grid-box is rejected. Minimum distance should be given in order to reduce the instability related to CFL condition in a river routing model.
Then, the potential outlet pixel with maximum upstream drainage area but not rejected in the previous step is selected for each grid, and the distance toward the selected potential outlet pixel of downstream grid is calculated. This step is repeated until the condition of minimum distance is satisfied in all grid-boxes.
When the condition of minimum distance is satisfied, the potential outlet pixels selected at the time are finally decided as outlet pixels. And, the elevation of the outlet GDBD pixel of each grid is decided as the channel elevation of the grid. Outlet pixels are marked with small green squares in the figure below.


River Network Map

Second, the fine-resolution river channel of GDBD was traced from the outlet pixel of a target cell until it reached the next outlet pixel downstream. The coarse-resolution cell where the next outlet pixel was located was determined to be the downstream cell of the target cell (the downstream cell of each cell is indicated with a blue vector in the figure below). Note that the downstream cell was not necessarily selected from the eight cells neighboring the target cell, but was located on any cell. The exception was if the traced river channel reached a coastal pixel of GDBD, then the target cell was recognized as a river mouth cell.


Channel Length and Distance to Downstream Grid

Third, the channel length of each cell was determined as the length of the fine-resolution river channel between the outlet pixels of the cell and its upstream cell (bold river channels within each unit-catchment in the figure below). Where multiple upstream cells exist for one cell, the upstream outlet pixel with largest drainage area was chosen as the representative upstream outlet pixel. For headwater cells with no upstream cell, channel length was given as a half-length of the cell size. Similarly, distance to a downstream grid point, X, was decided as the length of the fine-resolution river channel meandering between the outlet pixels of the cell and its downstream cells.

CONST_NETRORK_IMG
The constructed river network map. The downstream grid of each grid is indicated with a blue vector. Outlet pixels are marked with small green squares.

define_catchment.F

In define_catchment.F, unit-catchment area and floodplain elevation profile is decided.


Unit-catchment area

First, the group of GDBD pixels that drained into the outlet pixel of each cell was determined as the unit-catchment pixels of the cell (Pixels divided by black boundaries in the figure below). The total area of the unit-catchment pixels for each cell was determined as the catchment area, Ac, for the cell.

CONST_NETRORK_IMG
Unit-catchment for each grid.

Floodplain Elevation Profile

The cumulative distribution function (CDF) of the elevation within each unit-catchment was taken to describe the floodplain elevation profile (see the bold grey line in the figure below). This CDF describes the relationship between a flooded area and floodplain water depth. Every tenth percentile of CDF (dots in the figure) was extracted, and the line connecting these points (sharp line in the figure) was determined as a function to describe the floodplain elevation profile. The 10th percentile value is stored in 1st record, the 20th percentiles is stored in 2nd record, and so on until the 100th percentile is stored in 10th record of 'fldhgt.bin'.

CONST_NETRORK_IMG
Floodplain elevation profile.

combine_area.F

Combine continental maps to the global map

In combine_area.F, upscaled maps in '$FLOW/1deg/data/' for each continent is aggregated into global maps in '$FLOW/1deg/map/'.

set_map.F

Calculate supplemental maps

In set_map.F, supplemental maps related to the upscaled river network map are generated. Those maps are used for illustration in the next section, or simulation with rectangular grid box in CaMa-Flood instead of an unit-catchment approach.

Supplemental maps
NameDataFormat
rivseq.bin# of grids from far most headwaterInteger
grarea_grid.binRectangular Grid-box Area4-byte real
uparea.binUpstream drainage area (Unit-catchment based)4-byte real
uparea_grid.binUpstream drainage area (Rectangular grid-box based)4-byte real
nxtdxt_grid.binDistance between centers of upstream and downstream cells4-byte real
lonlat.binLongitude/Latitude of outlet pixels4-byte real
basin.binRiver Basin ID4-byte real
bsncol.binColor of River Basin for Illustrating4-byte real


station.F

GRDC gauging stations

In station.F, upstream grids of each GRDC gauging station is listed up. For example in the figure below, upstream grids of the gauging station (Light blue square) are decided as grids (b) and (d).

STATION_IMG
Upstream grids of a GRDC gauging station.


back to front page