There are quite a few factors that can influence minimum flow. To be reliable, usually this is done with streamflow data. Best to go oit during dry season or drought, and measure as many sites as you have time. One unusual example I remember was Dry Creek in coastal Oregon. It had the highest salmonid rearing habitat, but the channel substrate was very coarse and extensive sections were dry on the surface, with hyporheic flow in the substrate. Stevens Creek in S. Carolina andnthe Suwannee River in Florida are examples of streams with geologic faults, lineaments or karst substrate where sections have disappearing channel flow. Missouri had areas of karst topography with minor and major springs appearing with no special indicators up slope. My experience did not include much of HEC HMS, so I have not tried to use it for your desired purpose. I would most likely depend on available network of steam data or actual measurements during minimum flow, and other indicators such as sometimes culverts have a darkened area of low flow and aquatic macroinvertebrate indicators. Indicators of hydric soils along stream margins may also suggest permanence of saturation or flow.

As far as watershed delineation, there should be limited issues in hilly or mountainous terrain, but obviously more difficult in the coastal plain or flat lands. I have uploaded a couple of papers I co-authored on the advantages of LiDAR for georeferencing streams, hydrologic boundaries and picking up added details on past hydrologic modifications such as drainage ditches, beaver dams, etc. In my paper on Stream Types and Management Implications, I discussed the use of topographic crenulations and stream ordering and how it was helpful in estimating drainage boundaries, network, ordering (Strahler) and indicators that may help differentiate between perennial, intermittent and ephemeral channels.

William F. Hansen Thanks for your reply but I am looking for something different.

Hello Mr. Borse,

based on your description, I see no problem with your delineated basin. I even using HEC-HMS for estimating flows for smaller basins (around 0,3 sq. km) using only one basin without subbasins. It depends on the outcome of your work. For my Diploma thesis, I delineated one basin (approximately 2,5 sq. km) which I split to around 14 subbasins. The delineation was based on flow accumulation lines, which correlated with terrain survey and knowledge and experience of locals. In the Czech Republic we have a threshold of 0,3 sq. km for "critical points" which are considered as basins which can be vulnerable to the formation of flash floods. And for that basins analyses of that type are carried out (backed by official methodologies). If your basin does not contain visible stream network, so consider only ground runoff following NRCS Velocity Method for estimating the Time of Concentration.

Ondra Ulrich Thank you So much for your response. It will degenerately help me to develop a model on Small Scale.