top of page
  • Writer's pictureGeorge H. Allen

First published results from the LakeFlow algorithm

Updated: Jun 26, 2023

Effective water resource management depends on our ability to monitor and understand lake and reservoir inflows and outflows. Satellite remote sensing of lakes and rivers has become increasingly important for water management but little work has been done to estimate streamflow at river-lake interfaces. Riggs et al. (2023) introduces the LakeFlow algorithm, which uses SWOT observed river and lake variables to estimate inflow and outflow discharge. LakeFlow consists of a modified version of Manning’s equation from Durand et al. (2014) combined with lake volume estimates via mass balance. The algorithm applies Bayesian inference to constrain the uncertainty in LakeFlow’s unknown parameters given repeated SWOT observations (figure below).

Conceptual diagram of the LakeFlow algorithm which uses repeat SWOT observations of lakes and rivers to estimate the inflows and outflows of lakes in cubic meters per second.

Testing LakeFlow using synthetic SWOT data at three sample reservoirs (with 4 inflows and 3 outflows in total) yielded promising results: a median NSE = 0.88%, a median RMSE = 29.0%, and 5 out of 7 reaches having a bias < 15% (see figure below).

The image above shows the discharge estimated by LakeFlow for all lake inflows and outflows compared to gauge records.

LakeFlow could potentially be implemented on more than 16 thousand lakes and reservoirs worldwide (see figure below). As SWOT data becomes available in the coming months, we plan applying LakeFlow on global lakes and rivers, a step toward providing a more integrated understanding of surface water dynamics.

The image above shows the global distribution of lakes suitable for LakeFlow implementation (N = 16,610) with three sample lakes highlighted.


Riggs, R. M., Allen, G. H., Brinkerhoff, C. B., Sikder, M. S., & Wang, J. (2023). Turning lakes into river gauges using the LakeFlow algorithm. Geophysical Research Letters, 50, e2023GL103924.

George Allen is the head of the Global Rivers Group at Virginia Tech, Department of Geosciences.

34 views0 comments

Recent Posts

See All


Commenting has been turned off.
bottom of page