Today, more and more satellite data sets on rainfall, evapotranspiration and other hydrological variable are freely available from several Internet. These permits simple spatial water balancing exercises to be made. Sparse ground observations or data can be used to check or bias control the satellite estimates. This method is applicable to most areas of the globe. Of course, one needs to have some minimal scientific skills with these data processing, using open source or proprietary RS or Gis system soft wares. I hope this adds a bit to your expected answer.

Following the previous comment, you may have a look to this very recent paper:

"A worldwide analysis of spatiotemporal changes in water balance based evapotranspiration from 1982 to 2009"

http://onlinelibrary.wiley.com/doi/10.1002/2013JD020941/abstract

But as basic approach, you may have a look to the Thornthwaite-Mather method. See, for instance, the following documents:

http://www.sciencedirect.com/science/article/pii/0022169486901241

http://www.hydrology.bee.cornell.edu/Ecohydrology/Handouts/TM10.pdf

and here you can find a code:

http://wi.water.usgs.gov/Soil_Water_Balance/index.html

rainfall data and river flow data are the two basic parameters you require. There are methods which helps you to assess the water balance from just two of these variable (combined with some assumptions and using universal values).

PET can be estimated in various levels, with bare minimum data like temperature alone or using models (IWMI, awhere, world clim).

using GIS/RS technologies is another option to predict the water and temperature data, but as you suggested they need skilled persons and cost.

Thank you all very much for your suggestions and links; they are very helpful. Yes, it does seem true that some basic familiarity with accessing, downloading and working with global/regional climate datasets and simple models is necessary, followed by ground/field validation if possible.

We once developed (I think my colleague still works on it) a water management tool (SimSalin) for the Yanqi Basin in NW China which is about 180 x 80 km^2, and the model used Modflow from the USGS http://water.usgs.gov/ogw/modflow/ as basis. You can work on the basis of sattelite images and coarse geology, our model involved four geological layers and simulated five years on an office computer in some seconds. Of course, we had quite detailed data based on PhD thesises (http://e-collection.library.ethz.ch/eserv/eth:867/eth-867-02.pdf)

I recently tested TRMM version 7 precipitation data and temperature from CFSR reanalysis dataset (to compute potential evapotranspiration using Hargreaves-Samani method) to compute discharges using a hydrological model in the Mekong river catchment in South-East Asia. Both datasets area available for most of the globe and can be used in a hydrological model in areas with sparse ground based data. You can also use APHRODITE precipitation dataset, if your target area is in Asia

Please consider the use of easy and simple formulas like those by Budyko, Schreiber orTurc Pike among others. They give you an initial estimation of mean annual runoff depending on rainfall and pet.

Some references may be found in Singh, Hydrologic Systems: V. 1. Rainfall-runoff Modeling, Volumen 1 and watershed systems volume 2. Also a compilation in V. Arora articles

Arora's work in http://www.ec.gc.ca/ccmac-cccma/default.asp?lang=En&n=51CA77EB-1

Hi there,

Here you find a simple water balance model that requires basic data and its free for universities and developing countries.

http://weap21.org/

http://www.bgr.bund.de/DE/Themen/Wasser/Projekte/abgeschlossen/TZ/Acsad_dss/dss_fb.html

cheers

marcel

I have an Excel Worksheet

http://www.hydrology.uni-kiel.de/forschung/projekte/simpel/simpel-bodenwasserhaushaltsmodelle

But you can also program a simple model yourself with the spatial modelling syste PC Raster - we do it in a course.

Greetings,

Georg

Thanks again, everyone, there is a lot of usable and practical information here. I too used Excel for water balances, and Foxpro for storing daily data on rainfall, met parameters, inflows, water levels etc in a database, and programming to query them and do some math/stat ops on them. The question of simple water balance tools arose when i was talking with water resource managers and hydrologists in some African and Latin American countries, where some basic rainfall and discharge data exists, but no water budgets.

Hi there,

It seems, making the effective coordinates of water balace equation in your region as excel file, will be the best way.

Howdy,

In R there are a few nice tools. One that I am very familiar with, the EcoHydRology package, has a very simple water balance as well as a function to obtain 32 years (1979-2010) forecasted estimate of daily surface weather variables for any land location on earth (which often represents the weather as well as weather stations over 10km away from center of 40km basins). There is also a very simple "Spherical Cow" 9HRUs in 3subbasin SWAT initialization that can be easily set up, as well as a simplified initialization and call to the topmodel package. It also runs on any OS, so you are not limited to using MS Windows products (though you can) if you are a Mac/Linux user.

http://cran.r-project.org/web/packages/EcoHydRology/EcoHydRology.pdf

Temporal scale is important to know to chose the right tool. Are you interested in long-term water balance? If so, a simple ET model is all you need. An accurate ET estimate is essential for constructing a water balance for any region.

Thanks Daniel, an r function calling upon global datasets is very cool ! Thanks for the info.

Thank you, Ge, i'd start from a daily-monthly-annual water balance to get an idea of the relative magnitudes and seasonal variation of the individual balance components. Then go on to multi-year if data is available (at the very least, rainfall) although errors are likely to swell if one has to assume the same ET ( which means the same land use and weather conditions), similar surface_GW runoff/recharge partitioning and human water abstractions/discharges. I agree with your suggestion, of using a simple ET model (temp or radiation based like Hamon or Abtew) if its a long-term water balance, or one over a large spatial scale that averages out ET increases and decreases based upon vegetation type and season. Would you suggest any other ET model, if only say temperature or net radiation is available, and not RH, wind speed that leaves out using PM, PT models etc ?

Amartya, Seems to me any radiation based model is better than Temperature only model. Priestley Taylor is a good one. See this paper attached. For monthly ET, a first order estimate of ET for different landcover may be made with a equation in this paper if one has FAO PET data (i.e., local weather station data'..

http://admin.forestthreats.org/products/publications/A_general_predictive_model_for_estimating_.pdf

I suggest you trying WetSpa, a simple and parsimonious model, yet physical based that simulates hydrological processes of precipitation, snowmelt, interception, depression, surface runoff, infiltration, evapotranspiration, percolation, interflow, groundwater flow, etc. continuously both in time and space, for which the water and energy balance are maintained on each raster cell.

http://www.vub.ac.be/WetSpa/downloads/WetSpa_manual.pdf

Here is an application of WetSpa model:

Article WetSpa Model Application in the Distributed Model Intercompa...

Well, we have been working on large areas only in Tuscany and in the Volga river source area (3500 km2). The approach combined (1) hydrological modeling using all data available (including historical and present records, satelite, atc. with pixel size 4 ha) and (2) sap flow measurements in such pixels, verifying the modeled results, after up scaling data measured on series of trees (at least 12 per species and pixel). Such systems work automatically together with soil water and weather complexes and are very flexible considering space and time. Jan

Dear Dr. Cermák,

very interesting...

to which system are you referring? Can you provide a reference?

Kind regards,

L

WetSpa apparently is another model that can give full hydrologic balance if you have all the input data. This model apparently is not very explict about how landcover types affect ET. Like many other models, PET is estimated by the FAO method for grass. So, the question is: whould the assumption be ok for other land cover such as forest lands. I think it might be ok for arid regions since PET is much higher than actual ET. However, for wet area where AET of forests can be higher than referece ET (PET), such method might underestimate forest actual ET from the start.

A versatile tool to estimate real evapotranspiration and groundwater recharge is provided on

www.small-scale-hydrology.de.

Kind regards, K. Bohne

Thanks Alireza, Ge, Jan and Klaus. Its great to know of different models with different approaches and requirements, very useful for comparison too.

As regards actual ET, the problem still remains of how AET is estimated for different land cover types. For instance, in tropical mountains, there is a significant difference in canopy interception, infiltration and water uptake between native cloud forests with a thick interwoven canopy,monoculture forestry plantations that are more open, ecucalyptus plantations that have much higher water uptake rates than native forest. The FAO PM or equivalent vapo transport based models that most hydrological software uses is valid for alfalfa, as Ge mentions, and that too with the assumption of well-watered conditions. Hence its suitable for irrigated agriculture, but could be an overestimate for rangelands especially in the dry season, as well as be an underestimate of ET in forests in the wet season. I guess it also depends on scale of the water balance. In many watersheds, the approach of P = ET + runoff works as lomng as the watershed is closed (not much net GW or storage change) as well as the existence of accurate stream discharge data.

Schematic models are fine in homogeneous or more or less stable conditions and not much species. Otherwise you could get rather strange data when beleiving they are always right. One of the most simple approaches is application of suitable sap flow methods, which you may apply under any terrain and climatic conditions, low cost and easily movable equipment. Within-tree variation is important, but this is well known and problems may occur can be rather easily solved (see e.g. Trees 18:529, 2004) and others.

SWAT is a simpler model with limited inputs. Check it out

Hello,

There are different tools available to calculate the water balance of a watershed for different agro-climatic zone. More or less many water balance equations are more or less developed based on similar input/output relationship. The main issue in water balance analysis is how you estimate each component of hydrological cycles. That is the main point. How do you effectively calculate ET, how do you predict the ground water flow, how you calculate surface runoff, percolation and so on. The quality of rainfall data is also important for relatively accurate estimation.

Suppose if we have to find impact of a water balance by simplifying the approaches for

1. Groundwater flow

2. Runoff

3. ETp

4. Moisture retention

with very less input parameters as Rainfall and temperature alone

which method we can use to get idea about significant land parcels. We approaches we can use to case 1 to 4.

As going details into every aspect will although increase the accuracy, but will increase dependency on other parameters., which will not cater to SIMPLICITY of model.

Any suggestions?

1. You can use the experts evaluation of the unknown data.

2. You can develop a balance for large (not small) water management areas that are better provided by data from various observation posts and stations.

3. You can see this my expirience in the my prezentation on UNEP/UNECE/EU meeting (July 2014) with technology of the hydrological water balance development for the river basins: 

https://www2.unece.org/ehlm/platform/download/attachments/31522834/6_Mokin%26Greben_OSCE-WM-balance%20project7.pdf?version=1&modificationDate=1405930435016&api=v2

1. You can use the experts evaluation of the unknown data.

2. You can develop a balance for large (not small) water management areas that are better provided by data from various observation posts and stations.

3. You can see this my expirience in the my prezentation on UNEP/UNECE/EU meeting (July 2014) with technology of the hydrological water balance development for the river basins: 

https://www2.unece.org/ehlm/platform/download/attachments/31522834/6_Mokin%26Greben_OSCE-WM-balance%20project7.pdf?version=1&modificationDate=1405930435016&api=v2

See detail information about our methodology and tools to compute a basic water balance of any district of the river basin in the chapter 3 "The Methodology of the Build and Forecasting Balance of Water Management of District Basin River" (there are models, algorithms, maps and the real example for Southern Booh (Bug) river) of my monography (2014) (Ukrainian)

Book Scientific Basics of Rational Use Water of Ukraine by the Ba...

Hello,

I think the good method for your needs is Thornthwaite & Mather (1955) monthly water balance method. Simple method, based only on mean monthly temperature, monthly precipitation, geographical location of the study area and the available soil water content (AWC). For calculation you require mean temperature and precipitation, other parameters can be estimated indirectly. Also, in the original method temperature and geographical location is used to estimate Potential Evapotrasnpiration (PET), but in the water budget you can use any other (more precise) ET equation such as FAO 56 Penman-Montheith (calculating ETo for standard grass vegetation, ET limited or ETc for specific vegetation). One of the results of the water balance method is water surplus/runoff.

Thornthwaite & Mather (1955) monthly water balance method:

https://www.researchgate.net/publication/292150890_Thornthwaite_Mather_1955_The_Water_Balance_-_Multiple_Years_Worksheet

Software to calculate Reference evapotranspiration (ETo) according to FAO Penman-Monteith equation:

http://www.fao.org/nr/water/eto.html

Explanation of water balance method:

http://www.hydrology.bee.cornell.edu/BEE3710Handouts/TM2013.pdf

Hello again,

also, good method for your needs is the soil water budget according to the Wayne C. Palmer (1965). Method is widely used for calculation of the Palmer Drought Severity Index (PDSI and scPDSI), but in process all the elements of the water balance are calculated. 

Data Palmer (1965) Z-index and Water Balance - Multiple Years Worksheet

I found this very practical software for Thornthwaite monthly water-balance:

http://wwwbrr.cr.usgs.gov/projects/SW_MoWS/Thornthwaite.html

http://pubs.usgs.gov/of/2007/1088/pdf/of07-1088_508.pdf

Thanks Vitalii and Sabrija for the links and info from your experience. I'll go through them. These are definitely useful to obtain a first idea of the water balance in a river basin. I also found 1km-reolution MODIS_16 global ET datasets to be useful.

Now, smaller-scale studies of ET for different land cover ( vegetation ) types can help to narrow the uncertainty. Sapflux, diurnal groundwater change, eddy covariance ( energy budget), lysimeters are some experimental methods that can ideally be taken up by research institutions and universities in developing countries. One challenge is to increase the awareness amongst classically trained hydrologists ( in the engineering mould) about the tremendous variation in ET associated with different vegetation forms. Most river basin models lack that sensitivity, or add in vegetation components by coefficients whose values were determined by experimental studies in other regions, and hence may need to be determined for the region of interest.

try to use Wetspass model