depends on the size and scale of your study area.

upto 1:50000 scale landsat series images are good( having several bands) and can be downloaded by USGS website.

Sorry if I get wordy, but I think your question is important, and has ramifications for not only your specific circumstance, but any professional that is faced with a myriad of hydrology questions possible.. Unfortunately, I lost my original, more concise response as tablet locked up.  Before beginning, for some things, the 1:50K landsat is useful.  

Selecting data may depend on your hydrologic study topic and intent, perhaps even region or scale.  Remote sensing is seldom fully adequate, but very useful as visual tool and especially within a GIS format when scale can be adjusted.  My favorite remotely sensed data at this time is high resolution LiDAR.  As far as aerial photos, I think infrared is often best to highlight pine lands from hardwoods, and wet areas, streams from dry lands. Google Earth is pretty good for some things, and I am not sure if they use the 1:50K landscat, but it allows substantial scale changes, visualization of land use, basic topography, approximate stream location and type, etc.  Working in GIS ArcMap has much utility, especially if land uses have been classified and DEMs are available to evaluate topography, geology types and soil maps, vegetation types, road network, etc.  Since I have often worked in areas with dense canopy, these are my preference, but other areas may use other types of remote sensing to their advantage.  But remote sensing can only do so much relative to hydrology, as it would not typically yield a close look at the stream or river, or be suitable for measurements on-site, such as stream cross section, channel substrates, stage discharge measurements, identifying bankfull or flood indicators available, or provide data needed to validate model estimates.  

From tools such as high density LiDAR, we may detect improved hydrologic boundaries, geo-referenced stream networks, and historical detail or activities such as ditching or draining of wetlands that may not be noticed without intensive field evaluation when working in dense vegetation and flat terrain.   

Many times, a hydrology problem deserves awareness of the extent of applicable data and/or models specific to the problem being addressed.  An idea of precision needed may also be needed before making an informed decision of what data and process to select.  In some instances, this decision may not be our own, and we must notify others of limitations of existing information which may prevent achieving answers to the question or intent. Other data or model verification or adjustment may be needed for circumstance.  These decisions are not always easy to make, there may be several tools and methods of estimation or prediction, but also necessary sometimes to say that remote sensing and currently available information is not enough to meet the intent and precision desired.

My concern is that sometimes remote sensing and modeling may be selected and used when ground truthing and more intensive data collection and analysis is also needed and justified.  I think with time, we all gain the experience as professionals to make this call.  But unless we say or carefully document something, those that use our information or services are not aware of the assumptions, limits, precision, uncertainty involved.  When it comes to life and property decisions, we may not have had adequate training to factor hazard, risk and public safety needs into our technical analyses.  

I have indicated this at other researchgate responses, but as hydrologists or engineers, we may be prescribing things that involve life and property.  Whether doing at a University, agency, company, consultant or individual, we need to make sure we properly qualify results or add disclaimer when public and resource health are involved as to limitations, uncertainty, etc.  Non-professionals even with high positions of decision making are unlikely to be aware of the specifics of our work.  It is up to us to inform them.  

If we are doing things as a class exercise or general planning estimate, the need for precision may not be great, but my contention at the graduate, doctorate and professional levels, we need to be put through enough of an example, where we can begin to understand the intensity and importance of our advice, especially when associated with public safety and infrastructure design.  A couple years back, I saw that loss of life due to floods was one of the highest -- so more may be needed from us to help notify and address hazards.

It depends on your objective, size and scale of your study area.