It is possible to use satelite images or aerial photos of different dates... But the migration of a river, is not a linear precess, you need to determine the different events which can be at the origin of the migration (important floods...),

See this manuscript, it may help:

https://www.researchgate.net/publication/226832850_Meandering_and_bank_erosion_of_the_River_Nile_and_its_environmental_impact_on_the_area_between_Sohag_and_El-Minia_Egypt?ev=prf_pub

Article Meandering and bank erosion of the River Nile and its enviro...

Dear Nabajit,

talking about fluvial rivers & taking into consideration the time span, you need to think of migration rate. Yes I agree with Jean that good resolution satellite data/ aerial photographs can prove one of the good source along with the historical data which may be available from existing sources. You need lot to think about scale/positional accuracies. These are to be studies in consonance with factors causing migration and their frequency.

Good luck

I would agree on the term migration rate. The determination of the riverbed migration intensity comprises the horizontal movement distance between the first and last year of the observed period. This riverbed migration could be expressed in absolute terms. Therefore, the ratio of the old to new position could be determined for each river bank on the river profiles where the changes occurred. The total amount of riverbed migration was calculated as the average deviation of the maximum distances at all of these profiles on the left and right sides of the riverbed.

First of all I would like to Thank everyone for their responses. As far as data is concerned I am using Landsat images, IRS liss 3 images and topographical maps (SOI). The dataset is spread over 40 years 1970-2010. The banklines and the centreline (not the Thalweg) is delineated using Arc GIS and then the maximum deviation in each segment of the river is recorded. We added the values for each period and then divided it by time elapsed to determine the rate of lateral movement.

@ Richard Gloaguen: I am not sure about the directional preference but in order to calculate the rate we have nullified the direction and added the deviation for a particular period. Maximum lateral migration in each segment have been recorded for the study period.

@Ayman Ahmed: thank u for the link.

@Bhagwan Singh Chaudhary Sir: The region is annually affected by floods and the most significant cause for the shifts as we have found till now is the avulsions triggered by repeated breaches in the national highway-52 which intersect the rivers under study. The highest migration rate has been observed for the reaches downstream of the national highway intersection.

@Jean Albergel and Slavoljub Dragicevic: Thank you for sharing your insight on the subject.

Just a quick comment. Would not the migration of a river depend on its type? For example the migration of a braided river system (Brahmaputra, Ganges, Indus) or many rivers when they reach their deltas, be very different to a more geographically river say up near its source?

@Peter Webster: The rivers under study are meandering but the study reaches I am talking about is mostly straight.

These are the tributaries and sub-tributaries of Brahmaputra.

In case you need another braided system with multiple mouths have a look at the Irrawaddy or Ayeyarwady in Myanmar (Burma). See my cyclone Nargis paper.

I do respect all contributions, I fail to comprehend the term migration rate as there are several raters which can with some precision may be determine as erosion and weathering rate which may be generalized for a wide area. However, with same precision it cannot be said and generalised, therefore, some scholars call it migration tendency or frequency of rivers. say a river is observed to have migrated in 5 years. Point A it migrated a distance of 1 km at B 5 km at C 50 km at D simply 500 m, is it possible to all sum measured distance leaving migration distance in between ignored divide it by 4 and resultant value again by 5 finding x and justifying x as x km/m annual rate of of an specific river over two observation?

As a rough indicator of the degree to which

rivers can alter flow regime downstream, we calculate the Impounded Runoff index,

IR (expressed as a decimal or percentage) for major Rivers, as

per Batalla et al. (2004),

IR ¼ reservoir capacity=mean annual inflow

I also consider the term rate for the shifting of river courses is not really reflecting the condition as the phenomenon is a complex one, I would say a random shifting. a rate should be used for a more continuous change. so, the simplest model that could be tried probably a kind of stochastic model such as following poison processes or a more complex one such as random walk.

Understanding river migration involves a central question "What is the type of the river?" The migration patterns of big braided rivers like the Brahmaputra differs considerably from the meandering rivers like the Burhi Dihing, as observed in the upper reach of the Brahmaputra valley in Assam. For big rivers, bank line migration and migration of the median path of the river are quite different. Moreover, migration and avulsion should be treated differently if we are concerned with the 'controls' behind the river migration.

@Siddhartha Lahiri Sir: Shift in the river bank due to avulsive forces is restricted to only one segment of the river reach, in rest of the river its purely meander migration.

Sir is there any other way to analyse the trend of migration for a river?

It is already suggested by Hidayat Pawitan. It is sometimes becomes random and river channel in meandering rivers abruptly take different course and again meets its originally channel. However, meandering is not so stochastic in channel not carved out in base hard rock it and is gradual through bank erosion. But, rate of erosion on force or power of running water that seasonally varies and due stochastic event due to extreme rainfall and water poured in main channel or head water area its force become so intense that it is very difficult to predict the the outcome of this force. However, in normal course a point is reached whereby a small increase in river force/power may cut neck of the river meander and channel may become again become straight in that part and instead of lateral shift meandering shifts longitudinally. If you know flow, slope and bank erosion rate of a fluvial river, in that case you may with some success may calculate shifting rate of a meander keeping in mind seasonal changes in volume of flow etc. in normal course.

If the river migration is being studied in a tectonically active valley and the plano-temporal changes are compared in the GIS environment with the structural elements like blind faults, retrieved from the seismo-tectonic map published by the GSI (keeping in mind the present study area), trend of migration can be used to analyse many interesting issues relating the trend of Quaternary basin evolution. Moreover, a quick appraisal of the migration behaviour of the rivers in the immediate neighbourhood may help to develop certain useful models.

Hi Nabajit, surely you already have read Albert Einstein's (the father, not the son) theory of river migration from 1930 or so? And you surely know the booklet by Yalin & De Silva, 2001? If not let me know that I can help.

Thank u all. i would definitely try to incorporate the ideas from the discussion.

Hi Nabajit and all,

Many researchers (e.g. fluvial geomorphologists, geologists, hydrologists, sedimentologists, hydraulic engineers) have overlain serial historical sources of river bank positions (maps etc.) to reconstruct lateral channel migration rates across floodplains. ‘Migration’ is a term which is accepted by most communities. There are many data sources to allow river movements to be quantified over the last 100-150 years or so (see below for a typical list of sources available).

For methods and some cautionary comments, you might want to look at my review paper of a large range of methods for doing this: it’s entitled ‘The Measurement of River Bank Erosion and Lateral Channel Change: a Review’, published in Earth Surface Processes and Landforms, vol. 18 (9), pp. 777-821). DOI is: http://dx.doi.org/10.1002/esp.3290180905. I’ll try and attach a pdf copy of the publication here.

Typical data-sources to extract meander position dates from include sequences of satellite images, aerial photographs (including Google Earth), rectified ground photographs, early and modern surveys, published maps and surveyor’s notebooks. Some of these sources may need checking/calibrating for ‘cartometric accuracy’ first, by measuring the ACTUAL distances between pairs of points on the floodplain as they are today (in the field, or from modern maps or aerial photographs) and comparing these measurements to the same distances portrayed on old maps. Corners of buildings which are shown on the early maps can work well for this.

You are working in a part of the world where there is a rich, 150-year tradition of very early published field surveys of river channel positions and lateral channel change (e.g. see the classic papers by: Fergusson (1863) on the Ganges Delta; Shillingford (1895) on the Kusi River; Central Board of Irrigation, India (1947), on the Ramanga River). Full references are in my review paper in ESPL above (see Research Gate too).

Today, the channel superimposition method involves stacking in a GIS the river course plots for different dates – all at the same scale - then calculating any shifts in position by a 'spatial differencing' process. The differences in position of superimposed river courses can then be converted to meander migration rate. Note that river channel change rates can also be expressed in areal terms, if that is more convenient (e.g. in m2/m/yr), to assess the area of channel shift between any two banklines of known length over a given time-span. Professor John Lewin popularized this method in his excellent published work from the 1980s and 1990s.

For measuring contemporary erosion rates - and to quantify the magnitude, frequency and timing of individual bank erosion events with respect to floods, rainstorms, freeze-thaw episodes etc.) in order to provide a stronger basis for identifying the controlling erosion processes and resistances, you could look at the automated bank erosion monitoring techniques I have developed, called the PEEP and PEEP-TCT sensor systems (PEEP = Photo-Electronic Erosion Pin; TCT = Thermal Consonance Timing concept) (e.g. see Lawler (2005 and 2008) papers – also on Research Gate). They have been applied by ~30 research groups around the world now for erosion dynamics studies - not just in rivers, but also for gullies, beaches, tidal channels etc.

I hope these thoughts help! Good luck!

Picking up Helmut's point above, it is not widely known that Albert Einstein had strongly advised his son, Hans Albert Einstein, not to enter the field of river sediment transport because he thought it was too complicated (an interesting comment from a man who had developed the Special and General Theories of Relativity)! It was good for us that he ignored his famous father's advice, because Hans based himself in Berkeley and became a legend in the fluvial sediment transport field (e.g. see Einstein and Barbarossa 1950s papers etc.)!

One of the most accurate methods is using fallout radionuclides!

Please have a look on this paper:

http://www.sciencedirect.com/science/article/pii/S0169555X10003521

Many factors are possible responsible for the increasing trends, it is definitely because of increase in river discharge, rainfall events in the area or upstream, human induced also could be the cause.

@Helmut Baumert: I would definitely like to go through the papers u mentioned. If u could help it will be great.

@Damian Lawler: Thank you very much Sir for your detailed response. I am sure going to work on the points you mentioned and also try to explore the papers you suggested. Thanks again.

@Abdulvahed Khaledi Darvishan: Unfortunately the paper is not accessible from here.

Dear Nabajit

I think it's better to request the full text paper from author in this page:

https://www.researchgate.net/publication/223628343_Determining_lateral_migration_rates_of_meandering_rivers_using_fallout_radionuclides

Article Determining lateral migration rates of meandering rivers usi...

@Damian Lawler,

Your repose is quite interesting. Others have pointed methods and theoretical information, but the piece of information is highly valuable to me as I was never and I think most geomorphologists are not aware of it that Hans Albert Einstein was discouraged to enter the field of river sediment transport because he thought it was too complicated by his famous father, while CFD has become almost a specialized branch in domineering, maths etc. I have come across many responses of famous scientists when asked when you have opted one of social science course, why you did not opted for your major or in your higher studies. The most common response is nature is simple while society is complex. Of course, river sediment transportation is to a great degree is influence by human activity for one human use of land, however, in its basic form it is a natural process.

Thank you for this small but highly invaluable information.

Hi Nabajit,

I just met a website which might be of interest for you:

http://web.mit.edu/hml/ncfmf.html

There you can find something about secondary flows,

http://www.youtube.com/watch?v=yDw_7UIGCOo&list=PL0EC6527BE871ABA3

These phenomena are key for understanding of your river-mechanical topic.

But you surely have studied these things, only in case of. Enjoy,

Helmut

Thanks @ Helmut Baumert Sir

@ Juan Ballesteros-Canovas: Thank you for your valuable advice but I am a complete stranger to dendrochronology :-). Although I have read some literature, but truly, I found it quite difficult to grasp the subject and the methodologies.

Your first question was well answered by earlier posts, but I wanted to point you to a free add-on for ArcGIS. The National Center for Earth-Surface Dynamics (UMinn) has a Stream Restoration Toolbox on-line, and their Planform Statistics tool is quite useful for calculating migration rates (http://www.nced.umn.edu/content/stream-restoration-toolbox). There's an option to incorporate bend apex trajectories, and It can also draw a river centerline from bank lines and measure channel widths. It's worked well for me in the past, though I've run into some trouble using the apex trajectories when significant migration has occurred.

In response to your second question, I would recommend that you look into the factors that control sediment and water fluxes in your catchment. Have there been any changes in the catchment during that time period? Tectonic active in the headwaters or changes in land use that could have altered sediment input into the river? Do you have flow records? Have their been changes in flow regime, flood magnitude or intensity during that period? Is the river now flowing in an area with more easily erodible banks, or a more resistant bed?

Good luck with your study.

Bob

Yes, if you have dateable trees on the floodplain, then, as Hickin and Nanson’s work on the Beatton River, western Canada, demonstrated in the mid-1970s, you could complete a dendrochronological survey of the floodplain, and use that to estimate the rate of meander development. Their work was brilliantly innovative! Prof Gerald Nanson, now based in Wollongong University, Australia, has just received a lifetime achievement award from the British Society for Geomorphology!

Many workers have overlain in a GIS serial historical sources to reconstruct river or gully or delta or cliff/coastline movements over the last 150 years, but especially for river meanders (e.g. perhaps see my review paper of most methods of determining river migration rates in ‘The Measurement of River Bank Erosion and Lateral Channel Change: a Review’, published in: Earth Surface Processes and Landforms, vol. 18 (9), 777-821). DOI is: http://dx.doi.org/10.1002/esp.3290180905.)

I'll try and attach the paper here.

Other typical data-sources for previous meander position dates include: sequences of satellite images; aerial photographs (including Google Earth); ground photographs; early and modern surveys; published maps; and surveyor’s notebooks. Some of these sources may need 'cartometric' checking for accuracy before they are used in a superimposition model. Then the method simply involves stacking in a GIS the known river course plots for different dates at the same scale. The differences in position of superimposed river courses can be converted to meander migration rate, once all checks are completed.

@Robert Grabowski : Thanks for your suggestion and the tool although I am not able to find the tool even after installation, in my Arc GIS interface.

@Damian Lawler: I have gone through the paper and it is really of great help for choosing the method and type of data for channel migration studies.

If you have aerial photos of different dates with an appropriate scale for your sutdies, it is interesting a multitemporal analysis to chek the migration rates.

Nabajit,

I only just saw your follow-up question, so sorry for the delayed response. If you haven't already figured out how to use the Planform Statistics Tools, here is one way to install it:

1) Create a new toolbar by selecting Add-In manager in the Customize menu. Select Planform Tools and customize. Create a new toolbar.

2) Populate the new toolbar by opening up the Customize Mode window again (you can also do this by clicking on the down arrow in the new toolbar.) Select the commands tab, then Add-in controls, and the click and drag the commands to your new toolbar.

See the ArcGIS link for more info:

http://blogs.esri.com/esri/arcgis/2010/05/05/add-ins-for-arcgis-desktop-10/

Best of luck,

Bob