Spatial analysis in hydrology and water engineering

1
Spatial analysis in hydrology and water
engineering
Maa-123.420 GIS Analysis

• Ari Jolma
• 16.11.2004

2
Contents of the two lectures

• terrain analysis
• hydrology and spatial hydrology
• spatial aspects in water resources engineering

3
Terrain analysis

• Wilson and Gallant (eds.) Terrain Analysis
  (2000)
• Understanding the biophysical processes and
  patterns that occur at different scales in the
  landscape
• Most important data sources
• digital elevation models (DEMs)
• land-use classifications

4
Basic Topographical Attributes

• Altitude
• Slope (kaltevuus)
• gradient (steepest descent)
• Aspect
• slope azimuth (suuntakulma)

5
Calculation of slope and aspect

• Zevenbergen and Thorne (1987)
• Fit a 9-term quadratic polynomial
• z Ax2y2 Bx2y Cxy2 Dx2 Ey2 Fxy Gx
  Hy
• to the 8-neighbourhood
• assuming uniform grid, we get

6
Forces on the landscape

• Water
• Ice age, tectonics, wind, human, ...

7
Catchment
catchment divide / vedenjakaja
contour line / korkeuskäyrä
slope and aspect
40
channel network / uomaverkosto
Catchment outlet / luusua
8
Catchment shape
merging
divergent
rectangular
9
More topographic attributes and concepts

• Upslope and downslope area
• Specific catchment area
• upslope area per unit width of contour
• Flow path

10
Hydrological analysis of DEMs

• Assume a raster DEM
• Assume a neighbourhood of 8 pixels
• Local minimum
• Local maximum
• Flat area
• One lower pixel
• gt1 lower pixels

11
Flow direction raster

• Each pixel drains to zero, one or more neighbours
• Usual goal all pixels drain to exactly one of
  its neighbors, or outside of the DEM
• D8 algorithm
• cannot model flow divergence
• tends to produce flow in parallel lines which
  usually do not agree with aspect
• simple
• usable in catchment delineation
• Rho8, Dinf, FD8, FRho8 are other raster DEM ? FD
  raster algorithms

12
gt1 lower pixels

• in D8
• select the one to which the slope is steepest
  (distance between cornerwise pixels is times
  the distance between lateral pixels)
• in Rho8
• idea make a random change but take into account
  the slope
• in multiple flow direction methods
• divide the upslope area to all lower pixels
  depending on the slopw

13
Sinks in DEMs

• Real or error?
• To remove
• change DEM elevate pixels
• change flow direction raster
• drain to the lowest neighbour which does not
  drain to it
• drain to the lowest pixel just outside the
  depression

14
Flat areas

• Reason for the flatness?
• lakes (there is probably one outlet)
• fields (there is probably artificial underground
  drainage)
• Recursively drain to any pixel with drainage
  resolved or to the probable outlet(s)

15
Burning in and Breaching

• The river or outlet may be a narrow creek, too
  narrow for the DEM
• Burn in the blue lines
• locate the river pixels using some other
  information source and lower their elevation
• Breaching
• dig an outlet to the DEM using a breaching
  algorithm

16
Secondary topographic attributes

• Erosion indices
• potential erosion
• predicted erosion
• Potential solar radiation
• Indices for spatial hydrological models

17

• Elements of
• hydrology
• spatial hydrology
• hydrological engineering
• water resources engineering

18
Hydrological cycle
km3

• Picture from
• http//www.globalchange.umich.edu/globalchange1/cu
  rrent/labs/water_cycle/water_cycle.html

19
Floods

• A field in Rantsilassa 26.4.2000

20
Droughts

• Dying durra crop in Texas.

21
Water quality problems

• Algae blooms in Ekoln, Sweden

22
Reservoirs

• A new reservoir is needed(?) in Finnish Lapland
  to save water for energy production

23
Water transfer

• Spain is planning to use 25 G in a water
  transfer (from north to south) scheme

24
Dams

• China is building new dams in upstream Mekong -
  its effect in VietNam?

25
Climate change

• More winter floods and summer droughts to
  California?

Miller and his colleagues used the precipitation
and temperature data from the climate change
scenarios as input into the NWS "River Forecast
System," which is comprised of computer models
that can simulate river flow, soil moisture and
snowpack.
26
Spatial scales in hydrology

• Point
• Hillslope
• Catchment
• River basin
• Continent
• Global

27
Point

• How does a plant grow?

WSI 1 - (Actual Transpiration/Potential
Transpiration) multiplied by plant
susceptibility and integrated over the growth
season AT/PT f(soil moisture, weather
conditions)
28
Hillslope

• The path of water into a stream
• overland flow
• subsurface preferential flow
• soil water flow
• groundwater flow

29
ground-penetrating radar mapping
In this infrared image red regions indicate
vigorous vegetation, preferential flow pathways
are denoted with blue lines, high yield regions
are identified with thin black polygons, while
regions influenced by subsurface flow are shown
denoted by thick black polygons. Soil moisture
locations are represented with green stars.
Highest grain yields occurred where the
preferential flow pathways are within 2m of the
soil surface.
source http//hydrolab.arsusda.gov/ope3/Summaries
/fld_is001.htm
30
Spatial approaches to catchment hydrology
semi-distributed separate treatment for
hydrologically different areas geoinformatics
for preprocessing only
regular grid all processes treated separately in
each grid cell water routed from cell to cell
lumped spatial averages
31
River basin

• Large catchment, drains into sea

32
Spatial analysis regarding river basins

• Flooding
• Analysis of land-use changes in the catchment
• Analysis of flooded area
• Analysis of floodplains
• Siting of reservoirs and other structures
• Pollution transport (nitrogen runoff in the Rhine
  basin, Box 2.2 in Longley et al)
• Sediment transport

33
Soil Erosion and Sediment Transport in the
Yellow River, CHINAAmelia Chung
http//www-personal.umich.edu/acotel/Webpages/Cla
sswebpages/CEE590.html

• Amelia Chung
• CEE 590
• April 8, 2003

Picture Source Chinese National Geographic
Magazine, Vol. 1, 2001
34
YELLOW RIVER
35
The Legend

• Cradle of Chinese Civilization

• Food Basket of China

Picture Source http//yellowriver.org/huanghe
36
The Legend

• Chinas Sorrow

Picture Source O.J. Todd and S. Eliassen, "The
Yellow River Problem," Transactions American
Society of Civil Engineers, Vol. 105 (1940), p.
346. (Article published in December, 1938,
Proceedings)
37
The Geography

• Originates from the Bayankela Mountain in Tibet
  Highland, West of China
• Flows through 9 provinces, transverses 5,464 km,
  flows down from 4,500m to 400m
• Empties into Bohai Sea in Northeast China
• Divided into three reaches
• Upper, Middle, and Lower

38
The snow-clad Bayankala Mountains
Picture source YRCC, Huanghe Feng, Yellow River
Pub House, 1996.
39
Upper reaches
Lower reaches
Middle reaches
40
Soil Erosion

• The Loess Plateau
• Middle reaches
• Silt formation

Pictures source Land Resources in the Loess
Plateau of China, edited by The Northwest
Institute of Soil and Water Conservation,
Academia Sinica, pub by Shaanxi Science and
Technique Press, 1986.
41
Key Features of River

• Hydrology
• 745,000 km2 drainage area
• 47 x 109 m3 annual runoff rate
• 478 mm annual rainfall
• 1,200 mm evaporation rate

• Sediment
• Total 1.6 billion tons/year
• Average sediment concentration 35 kg/m3 (lt1
  kg/m3 in normal rivers)

42
Peak sediment load during flood flow, turning the
river violent
Picture source YRCC, Huanghe Feng, Yellow River
Pub House, 1996.
43
Sediment deposition in the riverbed near Middle
Reaches
Picture source YRCC, Picture Albums of the
Yellow River, China environmental Science Press,
1991.
44
Continental scale
45
The soviet plan to divert water south from big
rivers flowing north.
The fate of Aral Sea water was used for growing
cotton and did not reach the Sea.
46
(No Transcript)
47
Global scale

• The hydrological effects of
• ENSO (El Niño southern oscillation)
• NAO (North Atlantic Oscillation)

48
El Niño
http//www.cdc.noaa.gov/jjb/anim.html
49
El Niño, initially referred to a weak, warm
current appearing annually around Christmas time
along the coast of Ecuador and Peru and lasting
only a few weeks to a month or more. Every three
to seven years, an El Niño event may last for
many months, having significant economic and
atmospheric consequences worldwide. In the
tropical Pacific, trade winds generally drive the
surface waters westward. The surface water
becomes progressively warmer going westward
because of its longer exposure to solar heating.
El Niño is observed when the easterly trade winds
weaken, allowing warmer waters of the western
Pacific to migrate eastward and eventually reach
the South American Coast. The cool nutrient-rich
sea water normally found along the coast of Peru
is replaced by warmer water depleted of
nutrients, resulting in a dramatic reduction in
marine fish and plant life.
http//ww2010.atmos.uiuc.edu/(Gh)/guides/mtr/eln/d
ef.rxml
50
Atmospheric Circulation
Slide made by Thorsten Blenckner, Marko Järvinen
Gesa Weyhenmeyer
Sourcehttp//www.cpc.ncep.noaa.gov/products/preci
p/CWlink/pna/nao_loading.html
51
North Atlantic Oscillation
Slide made by Thorsten Blenckner, Marko Järvinen
Gesa Weyhenmeyer

• Impacts on local weather
• Temperature
• Snow fall
• West wind stress

Positive NAO warm winters Negative NAO cold
winters
52
The North Atlantic Oscillation (NAO) is a
phenomenon associated with winter fluctuations in
temperatures, rainfall and storminess over much
of Europe. When the NAO is 'positive', westerly
winds are stronger or more persistent, northern
Europe tends to be warmer and wetter than average
and southern Europe colder and drier. When the
NAO is 'negative', westerly winds are weaker or
less persistent, northern Europe is colder and
drier and southern Europe warmer and wetter than
average. One of the simplest definitions of the
NAO is that it is the winter difference in
pressure at sea-level between the Azores and
Iceland.
http//www.metoffice.gov.uk/research/seasonal/regi
onal/nao/
53
Hydrological models
54
Simulation models of hydrological processes
Evapo- transpiration
Snow storage
Melt / Sulanta
Soil water storage
to river
55
Types of models

• Conceptual models
• Lumped / separately for each subcatchment
• Semi-distributed model
• Separate treatment for each hydrologically
  similar area
• Grid-based model

56
Topmodel (Beven 1986)

• Relies on preprocessing of digital terrain data
  to calculate the catchment distribution function
  of topographic index
• a is cumulative upslope area (A) diveded by
  effectice contour length (L)
• beta is slope angle
• (weighted average if multiple flow directions)

57
SHE (Abbot et al 1986)

• Regular grid horizontal layers
• physically based representation of hydrological
  processes
• PDE / finite difference representation
• Empirical equations