Hello! Admittedly, "remote-sensing" is outside of my area of expertise. However, the answer to your question is "yes". Several examples of scientific research on this very subject are found below:

1) Assessing Surface Water Consumption Using Remotely-Sensed Groundwater, Evapotranspiration, and Precipitation (ANDERSON et al., 2012)

http://www.agu.org/journals/gl/gl1216/2012GL052400/2012GL052400.pdf

2) Integrated Methodology for Estimating Water Use in Mediterranean Agricultural Areas (ALEXANDRIDIS et al., 2009)

www.mdpi.com/2072-4292/1/3/445/pdf

3) Prediction of Crop Yield, Water Consumption, and Water Use Efficiency with a SVAT-Crop Growth Model Using Remotely Sensed Data on the North China Plain

(MO et al., 2005)

http://sourcedb.cas.cn/sourcedb_igsnrr_cas/zw/lw/200906/P020090625734725448762.pdf

Remote sensing alone can't give any conclusion without ground study

C. Hamilton's references are a great start. In general terms, you will determine what sort of remote sensing you wish to use: Landsat, MODIS, ASTER, AVIRIS are all good remote sensing imagery products, but have different spectral bands and pixel sizes. If using imagery, you'll need to do some ground-truthing to correlate spectral response with crop type, soil type, landcover type etc. On the water use side, using remote sensing to examine river levels and agricultural canal levels is tough--you are generally limited to measuring channel width from the images. You'll also need to control for precipitation vs canal use vs groundwater use.