What do you use in a runoff model if you do not have Edaphological (soil) data for an area? How would you simulate infiltration?
I am using MIKESHE by DHI to create a water balance model, more specifically a runoff model. In this software you add various data like daily precipitation, Temperature, snow melt, land uses from corine database (including root depth LAI etc) and you also add the catchment shape files and the DEM, and more. The problem comes when you need to know the field capacity of the soil, or the Permanent wilting point and the hydraulic conductivities - for every soil type defined from a shape file. But in my area there are no edaphological maps, so some people use geological maps and try to derive some soil classes. For example: limestone > hondrococcus soil > big permeability K = from ... to... .
Or use some hydrolithological maps that have the K ( hydraulic conductivity) ready to use.. but you wont know the Fc or PwP.
One idea is to make an overlay of the hydraulithological maps with the geological and use both so I can extract some info about the soil type. But there is another problem, what happens to the mountainous areas that there is no soil, like the bold mountain slopes? I cannot see these on my map except if I use Google Earth and then digitize the image with polygons. Too much. Or if use the slope and say for example that in slopes of 20% or above there can\t be any soil. Do you have any ideas?
Fieldwork is a must and if not possible choose another topic please! Sorry...
Aside from what Abbas suggested, you could consider looking into a topographic indexing method to figure out likely sources of runoff using something like the TOPMODEL developed by Beven. I mean, you're talking about a LOT of calibration with your proscribed methods...some physically based data based on your topo-index might be the way to go?
Hi, I can just tell you that you should not use such model, as you observed yourself, it is no realistic to collect such datas. Furthermore, the complexity of soil make also not realistic to use soil parameters on large scale catchment. Depending on the basin scale, you can use only geological datas (expressed in term of substratum permeability) as I made in those publications:
http://www.researchgate.net/publication/26640230_Assessing_winter_storm_flow_generation_by_means_of_permeability_of_the_lithology_and_dominating_runoff_production_processes
Article Assessing winter storm flow generation by means of permeabil...
you asking for urine without a kidney! i advise you collaborate with a pedology to at least get the major soil type of the area,then you can proceed with your work. All those short cuts you are looking for will not take you any further.
Soil data are very important in soil erosion study. Developing a runoff model model without the consideration of edaphic factor on which the runoff will act upon is not right. You can carryout some preliminary studies on the soil in your target area for it to give you an idea before proceeding to your model
It would be better to use inverse modelling alternative.i should mention that you can not say there is no soil in mountainous area.
Even if there are no edaphorlogical maps on ground, there should be some data on the intended soil in your university library or other universities around. If not (though i doubt), collect soil samples and analyse.
Hi, It could be good attempt for preparing runoff modeling, but i suggest you should take at least soil type and general soil chemistry data then you will get a fantastic model for further research work.
Dear, Mohini Singh, are you sure that soil type and soil chemistry will get information on the runoff? .I doubt
Soil type just give a name but doesn't get information on the hydropedological behavior of the soil.
Georges, I am soil scientist working with hydrologist, and I can tell you that what ask you Mikeshe in term of soil datas is unrealistic. To really feed the model with those soil datas, you need not only a very goog soil map (good spatial assement of the soil units) but also a lot of analyses to determine Permanent wilting point and the hydraulic conductivities of each soils horizons.
Finaly, the only solution is to play with the model and maybe you will obtain good results but probably for wrong reasons
Thank you all , i know that it is stupid to ask about a side way to do this. and i know the importance of soil field tests. .but the area is αλμοστ 1000 (square Km) so for the purposes of a Msc thesis it is impossible to do soil tests.Joy, In Greece we have some digitized or not , soil maps but mostly for agricultural areas .
Jamal i know that mountainous areas have soil but i was asking what could i do for the mountain slopes that are simple rock and i cannot see them in a geological map .I do not have a lot of time , and o think i am going to play with the model and calibrate it using the runoff data i have for the main River . If anyone has a not so another idea on making this better with no soil data i will accept it ! i wish i could change the subject this projects should be done by companies not by one student only . The place is Arta in Epirus Greece, upper stream of River Aracthos, till the dam .If anyone is wondering ! thank you all !
Since you don't have soil field test data you are going to have a pretty large margin of error, you can mitigate that by breaking the area you are studying into discrete regions (I suggest doing so by watershed) and using the average for those regions.
Hi, it is very braod scale but, given the size of your catchment, you could consider using the Soil Map of Europe - this also comes with some soil physical data (partly derived from the SPADE- Soil Profile Analytical Database for Europe). Also the Harmonised World Soil Database has soil physical data that can be linked to the Soil Map of Europe.
You are trying to model without knowing the location of the experiment? You are making the model by remote sensing? Without knowing the type of soil has no way to make the model accurately. Geology you can get an idea about the infiltration capacity, but errors may occur, and errors occur, your model is inefficient.
You should write to someone who knows the area and learn about soil permeability that occurs there.
i have 6 years of Runoff Peter , i will calibrate but there are also other parameters that need calibrations.. i know this model wont be accurate, but i will do my best , and after all its for educational purpose , Thank you all. Does MikeShe auto calibrates ? and specifically on a parameter i will choose? like soil? or do i have to try different values till it works? thanks!!
Hi, George. I know the greek problem to find data. The best solution for cases like this is the old but good model of Thorthwaite and Mather. I am sending you three articles two in greek and one in english based on the previous in order to undestand the method.
The thesis is finished now i uploaded on my page but it is in Greek .. the abstract is in English . and some diagrams are shown only if you download the document , i do not know why this is happening .. but if someone whats to take a look its better if he or she downloads it all. Thanks again.Edaphologically we did a good work amd simulated the base flow and interflow quite well .
- Badges
- Science topic
- Similar topics
- Agriculture
- Soil Science