Report on Advanced Engineering for Global Environment and Measurement
Topic:
River hydraulic measurements using
remote sensing
Prepared
for Prof RIKIMARU Atsushi,
Nagaoka University of Technology
Prepared
by Ayurzana Badarch 14701491
January 5, 2015
INTRODUCTION
This report is provided information about utilization remote
sensing technology of river hydraulics measurements and modeling. Remote
sensing technology is rapidly and widely used many fields especially issue of
environmental protection and its quality monitoring since taking remote sensing
advantages and capabilities. On the situ measurement is important but it has
some own disadvantages. In the some case, it leads to use remote sensing
technology for measurements. In this report, we consider that river discharge,
velocity, depth, water surface slope, bathymetry characteristics and some water
quality assessment which are important measurements to do accurate modeling and
to assess current condition of river environment. Many related literatures are
studied and published.
REMOTE SENSING FOR
RIVERS
Remote sensing of rivers, the topic of this special issue,
is rapidly developing as a new subdiscipline in the river science. Rivers are
continuous systems that vary across multiple space and time scales. To truly document
the range of river structures and functions therefore requires continuous data
across a wide range of spatial and temporal scales (Marcus & Fonstad, 2010) . Methods classically
used to map rivers, such as cross section or detailed reach scale surveys,
capture only a small portion of a river and often do not portray the range of
variations throughout the system. Remote sensing can provide continuous
coverage at varying resolutions on a repeat basis, thus creating the potential
to document a remarkable range of variations in river parameters (Aberle, 2011) .
ABOUT HYDRAULIC MEASUREMENTS
General hydraulic measurements are performed using permanent
gauging station which has been established on stretch of river where there is a
stable relationship between stage (water level or depth) and discharge, and
this has been measured and recorded certain time (E.Kuusisto, 1996) . Permanently measured parameters are
meanly water depth, velocity, temperature and sediment, they are changed over
time. Then using water depth and velocity is used to estimate discharge which
can be recorded. Other constantly parameters for over time are vegetation and
roughness which are important to modeling river flow, inundation of flood and
to evaluate some environmental condition. Above mentioned parameters and
measurements can fluently measure at gauging station, but considering among
river length, it is restricted by number of gauging station and its position.
This may be another reason to use remote sensing technology for hydraulic
measurements. Following sections are delivered glimpse of above mentioned
parameters and measurements excepting
temperature and sediment using remote sensing technology.
RIVER DEPTH AND BED
ELEVATION
Remote sensing of water depths dates at least to World War
II, when photogrammetric techniques were used with aerial photos to measure
near-shore depths in the pacific. Likewise, some of the earlier work on use of
digital imagery addressed techniques for estimating depth.
The Surface Water and
Ocean Topography (SWOT) radar interferometer satellite mission will provide
unprecedented global measurements of water surface elevation for inland water
bodies. However, like most remote sensing technologies SWOT will not observe
river channel bathymetry below the lowest observed water surface, thus limiting
its value for estimating river depth and discharge. Therefore (Mersel, 2013) explored if remotely
sensed observation of river inundation width and depth alone, when accumulated
over time, may be used to estimate this unmeasurable flow depth. They concluded
that their findings have positive implications for SWOT and other sensors
attempting to estimate river flow depth and discharge solely from incomplete,
remotely sensed hydraulic variables and suggest that useful depth retrievals
can be obtained within the spatial and temporal constraints of satellite
observations (Mersel, 2013) .
Figure 1. Bathymetric map of the Lamar River, WY,
generated using the HAB-1 technique with PC bands derived from Probe-1 128-band
imagery. Greater depths are indicated by darker blues. Flow direction is from
the bottom of the image toward the top (Fonstad and Marcus, 2005).
Da =(Q/(3.125WS0.12))0.55 (1)
Where Da, average depth,
can be estimated based on ground measurement of discharge, slope measurements
from maps, and with measurements from imagery.
The HAB technique does not require ground-truth depth
information at the time of flight. They presented two version of techniquie,
HAB-1 is based on geometry, discharge and velocity relationships of river
channel, HAB-2 is similarly of operation, but the assumption that the
distribution of depths approximates that of a triangle is replaced by optical
Beer-Lambert law of light absorbance. Depth maps and cross sections derived
from HAB techniques are consistent with typical stream geomorphology patterns
and provide far greater spatial coverage and detail than could be achieved with
ground-based survey techniques (Fonstad, 2005) .
Bathymetry, is
indicated river bed elevation, it can be basic data to modeling river
hydrodynamics. Westway et al, (2001) concerned with the technical aspects of
using hig-resolution digital photogrammetry and image analysis techniquies to
generate dense and accurate digital elevation models (DEMs). They had two main
aims: (1) assessment of the representation of exposed areas using large-scale,
airborne imagery and conventional digital photogrammetry; and (2) assessment of
the representation of submerged areas using same imagery, using a two-media
photogrammetric refraction correction model and image analysis techniques. In
the result of study, they determined that digital photogrammetry can be used to
obtain accurate high-resolution topographic information in certain fluvial
environments, despite the relative low relief and presence of water. The
quality of DEMs produced seems to be critically related to depth of water.
Where there is no water or water is very shallow and clearly (less than about
0.2m), the mean errors associated with raw photogrammetric DEMs are low (mean
error of 0.01 to 0.05m; standard deviation error of 0.04 to 0.01m). As water
depth increases, both the ME and SDE tend to increase. Study of (Westaway, 2001) has shown that
digital photogrammetry, when used with a two-media refraction correction, is
capable of doing this for clear-water, shallow, gravel-bed rivers.
Figure 2. The red dots provide an example of the
irregular spatial distributions of the CHARTS point data collections. The
underlying hypsographic shaded terrain surface is the resulting product (Coleman,
2010) .
The U.S. Army Corps of engineers joint Airborne LiDAR
Bathymetry Technical Center of Expertise (JALBTCX) operates a special Light
Detection and Ranging (LiDAR) instrument that is capable of penetrating through
water (up to 1.5x secchi disk depth) to collect both topographic and bathymetry
elevation data. This is called Scanning Hydrographic Operational Airborne LiDAR
Survey (SHOALS). In this surveying, deep water areas and near-shore shallow
areas are missing data. Next, in 2003, the U.S. Army Corps of Engineers joint
JALBTCX was tasked to use a next-generation bathymetric LiDAR technology,
referred to as CHARTS, to collect high-resolution bathymetric and topographic
data for areas adjacent in extent to the original SHOALS data collection area.
The CHARTS system is a highly highly-specialized system which integrates a
1,000 Hz hydrographic LiDAR instrument, a 10,000 Hz topographic LiDAR, and a 1
Hz digital camera. The lower-frequency component of the CHARTS system is
capable of penetrating water between 0.1 meters and 50-meters of depth (3x
secchi disk depth) and the average horizontal spacing for hydro-based points is
2-5 meters where terrestrial points are spaced at 1-2 meters. In the CHARTS
surveying, presence of riparian vegetation zones that needed to be cleaned or
filtered. Those LiDAR technologies are widely used in bathymetry surveying in
US (Coleman, 2010) .
RIVER DISCHARGE
The flow rate or discharge (main measurement) of a river is
the volume of water flowing through a cross section in unit of time and is
usually expressed as m3s-1. It is calculated as the
product of average velocity and cross section area but is affected by water
depth, alignment of channel, gradient (slope) and roughness of river bed. Originally,
discharge may be estimated by the slope-area method, using these factors in one
of variations of Chezy equation (E.Kuusisto, 1996) and also can be calculated simplest and
with several variations Manning equation.
The development of methods to estimate the discharge of
rivers using remotely sensed data would provide that means to increase the
stream flow measurement network globally (Bjerklie D. M., 2005) Remote sensing river
discharge has the potential to provide some needed data by filling in gaps
within the existing gauging station and by adding new information from
inaccessible regions that have not been gaged in the past. The use of remotely
sensed information to track changes in river discharge has been shown to be
feasible and potentially useful where ground-based (situ or gauging station)
data is difficult to obtain ( (Kuprianov, 1993) (Koblinsky,
Clarke, Brenner, & Frey, 1993) . For this reason,
site specific discharge ratings developed from ground-based flow measurements
and remotely sensed hydraulic information are not practical unless the
discharge ratings are transferable to areas where ground measurements of flow
are not available.
Estimating discharge in rivers from hydraulic information
obtained solely from aerial and satellite platforms has been explored and
summarized by (Smith, 1996)
and (Bjerklie, 2003) . (Bjerklie, 2003) has suggested that hydraulic
information data can be used to estimate in-bank river discharge using various
general hydraulic equations. In the paper of
them, water surface width and maximum channel width measured from 26 aerial and
digital orthophotos of 17 single channel rivers and 41 SAR images of three
braided rivers were coupled with channel slope data obtained from topographical
maps to estimate the discharge. (Bjerklie, 2003) used multiple resistance equation
similar to Manning equation that uses observable river channel hydraulic
information to estimate in-bank discharge in rivers. Then they suggested that a
general form for natural rivers discharge be defined as following equations:
Model 1: Q=k1WY1.67S0.33 (2)
Model 2: Q=K2WV2.5S-0.5 (3)
Model 3: Q=k3W1.67V1.67 (4)
Where W is the water surface width (m), Y is the average
water depth (m), V is average water velocity (m s-1), and S is the
channel slope measured from 1:24000 scale topographical maps and with k1,
k2, and k3 is representing a general conductance
coefficient which can be determined from large database of observed flow
measurements.
Figure 3. This figure shows an image of the Missouri
River near Elk Point, SD as collected by JPL’s AirSAR at C-band using its along
track interferometric capability to measure radial surface velocities. The
image shows the measured velocities after being projected into the horizontal
plane and corrected for the phase-speed of the Bragg-resonant waves. The radar
viewing orientation was South with the aircraft flight direction from east to
west (Bjerklie D.
M., 2005) .
Fig.3 shows the inferred horizontal velocity of flow and
regions of river that were used to obtain four discharge estimates by (Bjerklie D. M., 2005) .
The equations developed by (Bjerklie D. L., 2003) indicate that
discharge estimating models that include with, depth and slope have generally
greater accuracy, especially for larger rivers, compared to models that use
width and slope only or width, slope and velocity. In the result of research,
the standard error of discharge estimates were within a factor of 1.5-2
(50-100%) of observed, with the mean estimate accuracy within 10%. This level
of accuracy was achieved using calibration functions developed from observed
discharge. Without using a calibration function, the estimate accuracy was +72%
of the observed discharge, which is within the expected range of uncertainty
for the method. However, using the observed velocity to calibrate the initial
estimate improved the estimate accuracy to within +10% of the observed.
Remotely sensed discharge estimates with accuracies reported in this paper
could be useful for regional or continental scale hydrologic studies, or in
regions where ground-based data is lacking (Bjerklie D. M., 2005) .
Also (Colin J., 2013) tried to show that useful estimates of
absolute river discharge may be derived solely from satellite images, with no
ground-based or a priori information whatsoever. The approach works owing to
discovery of a characteristic scaling law uniquely fundamental to natural
rivers, termed river’s at-many-stations hydraulic geometry (acronym within
paper is AMHG). A first demonstration using Landsat Thematic Mapper images over
three rivers in the US, Canada, and china yields absolute discharges agreeing
to within 20-30% of traditional in situ gauging station measurements and good
tracking of flow changes over time (Colin J., 2013) .
Figure 4. Daily time series of observed river
discharge, nowcast (current) and forecast (for selected lead time) based on the
river flow signal observed from satellite (Hirpa,
2012) .
Q=I1/2KW(Z-Zb)5/3 (5)
Where I is water surface slope, Z is water level and other
terms similar to equation (2)-(4).
Proposed method was developed and tested primarily on data
prom two Amazon gauging stations and on simulated data and relative error in
the discharge estimates was fewer than 10% (Nergel, 2011) . Also they applied the five statistical
models used remote sensing approach including Bjerklies model. (Bjerklie D. , 2007) , who also studied
method to estimate the bankfull velocity and discharge in river that uses the
morphological variables of the river channel, including bankfull width, channel
slope, and meander length was developed and test. Because these variables can
be measured remotely from topographic and river alignment information derived
from aerial photos and satellite imagery, it is possible that the bankfull
state flow can be estimated for rivers entirely from remote-sensed information.
Definition of bankfull hydraulics of rivers would provide a baseline condition
for quantification of large scale regional changes in river morphology through
remote tracking of variables including with, stage, and slope, and serve as a
reference for remote tracking of hydraulic dynamics. He proposed several
regression equations to estimate bankfull velocity and discharge and some are
outlined as:
With most good correlation velocity equation is V=1.37I0.31lamda0.32,
r2=0.95 (6)
With most good correlation discharge equation is Q=0.24W1.64,
r2=0.90 (7)
In the result of study, he concluded that bankfull discharge
can be estimated remotely with a mean uncertainty on the order of 24% or less
for a large number of estimates, and that improvement in estimates of the depth
will reduce both the mean and standard deviation of uncertainty (Bjerklie D. , 2007) .
RIVER SURFACE VELOCITY
Essentially two non-contact methods have been used to
measure the water velocity at the river surface. The first relies on digital
images in visible or infrared (IR) that can be taken using video or picture
cameras such as Particle Image Velocimetry (PIV). It can be considered as a
passive detection methods since the natural emission of the river surface is
used. The second relies on radar detection using wavelength from microwaves to
radio waves. Sometimes, those are termed Local Remote sensing. Also to measure
the water velocity, other LiDAR or SODAR technology can be used but not
commonly used in this case (Creutin, 2001) .
BRIEF ABOUT ROUGHNESS
AND VEGETATION SURVEY
Floodplain roughness parameterization is one is one of the
key elements of hydrodynamic modeling of river flow, which is directly linked
to safety level estimation of lowland fluvial areas. Necessary input parameters
are median grain size for unvegetated areas, vegetation density for forest and
vegetation height and density for herbaceous vegetation. (Straatsma, 2006) presented a method for spatially
distributed roughness parameterization, in the entire floodplain by fusion of CASI
(Compact Airborne Spectral Imager) multispectral data with airborne laser
scanning (ALS) data. The method consists of two stages: (1) image segmentation
of the fused dataset and classification into the most important land cover classes
(overall accuracy=81 percent, and (2) determination of hydrodynamic surface
characteristics for each class separately (Straatsma, 2006)
For detailed hydraulic modeling, accurate spatial
information of roughness pattern needs to be derived in automatic fashion. (Forzieri, 2010) proposed a
supervised classification for heterogeneous riparian corridors with a low
number of spectrally separate classes using data fusion of a Quickbird image
and LiDAR data. The approach considers nine land cover classes including three
woody riparian species, brush, cultivated areas, grassland, urban
infrastructures, bare soils and water (Forzieri, 2010) .
IN MY PREVIOUS STUDY
I am one of valueless customer of Earth Resources
Observation and Science Center (EROS) branch center of U.S Geological Survey
(USGS). EROS is a remotely sensed data management, systems development and
research field center and provides with imagery and tool to explore Earth
science. I had been working on two projects using DEM data with 30m resolution
which is developed from Landsat 8 imagery. DEM is used to visualize flood plain
(how much area under water during the inundation) areas and to provide river
geometric information for approximated modeling using HEC-GeoRAS extension tool
of Arc-GIS.
Figure 5. Back water analysis with two cases;
without dam and with dam. Blue color indicated 100yr flood inundation without
dam and pink boundary indicated 100yr flood inundation and back water influence
with dam on the river. (Nasanbayar,
2014)
Also DEM is commonly used to delineate watershed, calculate
watershed characteristics and evaluating land uses for hydrologic modeling.
DISCUSSION
In open channel flow, river flow, key variable is discharge;
it can be calculated following equation of Manning’s
Q=(1/n)AR2/3I1/2 (8)
Where A and R is area of cross section and hydraulic radius that
is meanly assumed equal to half of depth of river in open channel flow. In
equation (7), other parameters excepting n depends of river channel geometry
and can be expressed water surface width and depth which are available to
remotely sense. I understood that utilization of remote sensing in river
measurements and monitoring is bigger than before I supposed.
Variety measuring and estimating methods using local or
space remote sensing technology are already proposed and tested, and achieved
to its purposes. Discharge measurement is based on river surface width, water
depth, slope, and of some times average velocity. Correct measurements of those
basic geometric and kinematic values lead to more accurate discharge
measurements. In those values average velocity is relative one which can be
evaluated using measurements of surface velocity of river. Comparing with
utilization of remote sensing on water quality measurement and monitoring,
river measurement has cogent deficient development.
According to statistics of 2006, it is not sufficient and
emerged issue in Mongolia that only 110 gauging station on the 74 rivers. There
are 4000 rivers with important 300 rivers (evaluated by priority) in Mongolia
used in water resources for wild animal and nomadic life. We really need study
to using remote sensing existing and developed capabilities for those unstudied
rivers to assess condition and monitor.
CONCLUSION
The growing availability of multi-temporal satellite data
has increased opportunities for monitoring large rivers from space. A variety
of passive and active sensors operating in the visible and microwave range are
currently operating, or planned, which can estimate inundation area and
delineate flood boundaries. Radar altimeters show great promise for directly
measuring stage variation in large rivers.
Researcher and engineers are exploring the use of airborne
remote sensing as a cost-effective way to gathering information needed for
river assessment and modeling in developed countries. In the Mongolia,
researcher use to reflectance information from Landsat imagery in order to
study only river quality measurements. Compared with lakes, rivers and streams
pose a challenging set of problems for application of remote sensing techniques
to river measurement either water quality assessment because:
1.
They are temporally more dynamic.
2.
The resolution of Landsat (30m) is too coarse
for rivers and streams.
3.
For researcher, better set of spectral bands
than the Landsat bands is needed or Mongolian’s own launched satellite.
We have may another solution which has been to use airborne
high-resolution hyperspectral imagery obtained from small aircraft flying over
stretched of rivers. I saw similar example in master research work of master
student of NUT, they are using mini helicopter to determine ocean surface wave
direction and magnitudes.
My current Ph.D study is not directly related to remote
sensing technology. Thesis is Ice formation process modeling using Smoothed
Particle hydrodynamics method.
Bibliography
Aberle, J. (2011). Use of remote sensing data in
environmental hydraulics. Norwegian University of Science and Technology.use of remote sensing data in environmental hydraulics
Bjerklie, D. (2007). Estimating the bankfull
velocity and discharge for rivers using remotely sensed river morphology
information. Journal of hydrology, 341, 144-155.
Bjerklie, D. L. (2003). Evaluating the potential for
measuring river discharge from space. Journal of hydrology, 17-38. http://www.sciencedirect.com/science/article/pii/S002216940300129X
Bjerklie, D. M. (2005). Estimating discharge in
rivers using remotely sensed hydraulics information. Journal of hydrology
309, 191-209.
Coleman, A. L. (2010). Development of
High-Resolution Bathymetry datasets for the Coulmbia River through the
Hanford Reach. Technical report of PNNL.
Colin J., S. L. (2013). Towards global mapping of
river discharge using satellite images and at many stations hydraulic
geometry. PNAS Volume 111, 4788-4791.
Creutin, J. (2001). Local Remote sensing of
rivers. Iowa City: Technical report of Iowa Institute of Hydraulics
Research.
E.Kuusisto. (1996). Hydrologic measurements. In E.Kuusisto,
Water quality monitoring – A Practical guide to the design and
implementation of fresh water quality studies and Monitoring Programmes.
United Nations Environment Programme and the World Health Organization.
Fonstad, M. M. (2005). Remote sensing of stream
depths with hydraulilly assisted bathymetry models. Geomorphology 72,
320-339.
Forzieri, G. M. (2010). Riparian Vegetation mapping
for hydraulics rougness estimation using very high remote sensing data
fusion. Journal of Hydraulic Engineering, Volume 136, 855-867.
Hirpa, F. H. (2012). Upstream satellite remote
sensing for river discharge forecasting: Application to major rivers in South
Asia. Remote Sensing Environment 131, 140-151.
Koblinsky, C., Clarke, R., Brenner, A., & Frey,
H. (1993). Measurement of river level with satellite altimetry. Water
Resource. Res. 29, 1839-1848.
Kuprianov, V. (1993). Aerial methods of measuring
river flow. In Aerial methods of measuring river flow: Herschy (Ed.),
Hydrometry: Principles and Practices (pp. 473-478). Wiley.
Marcus, A., & Fonstad, M. (2010). Remote sensing
of Rivers: the emergence of subdiscipline in the river sciences. arth
Surfece process and Landforms Volume 35, 1867-1872.
Mersel, M. S. (2013). Estimation of river depth from
remotely sensed hydraulic relationships. Water resources research, 49,
3165-3179.
Nasanbayar, N. A. (2014). Eg HPP preliminary
feasibility study. Ulaanbaatar: MONHEC research center.
Nergel, J. K. (2011). Estimating river discharge
from earth observation measurements of river surface hydraulic variables. Hydrology
and Earth system Sciences 15, 2049-2058.
Smith, L. I. (1996). Estimation of discharge from
three braided rivers using synthetic aperture radar satellite imagery. Journal
of water resources, 2021-2034.
Straatsma, M. (2006). Floodplain rougness mapping
synergy: LiDAR and Spectral remote sensing. Utrecht University.
Tarpanelli, A., Brocca, L., Lacava, T., Faruolo, M.,
Melone, F., Moramarco, T., et al. (2011). River discharge estimation through
MODIS data. Remote Sensing for Agriculture, Ecosystems, and Hydrology XIII
(p. 8). Antonino Maltese Prague: SPIE Proceedings.
Westaway, R. L. (2001). Remote sensing of
Clear-water, shallow, gravel-bed rivers using digital photogrammetry. Photogrammetric
Engineering and Remote sensing 67, 1271-1281.
No comments:
Post a Comment