Decide how to make use of mean depths, where available

[hcorson-dosch-usgs]

For lakes for which we do not have hypsometry data, but do have depth data, we currently use maximum lake depth. However, we do also have mean depths for some of those lakes within these data files:

• '2_crosswalk_munge/out/navico_nhdhr_xwalk.rds.ind'
• '3_params_fetch/out/LAGOS_NE_depths.csv.ind'
• '3_params_fetch/in/WI_Waterbodies.tsv.ind'

We could make use of those mean depths to get a better estimate of lake hypsometry, using something like the approx.bathy() function in the rLakeAnalyzer package, with the voldev method, which uses max depth, lake surface area, and mean depth. From the documentation for that function:

Estimates a depth-area curve for a lake using lake surface area, maximum depth and mean depth. Two methods for estimating the curve are available; ’cone’ assumes the lake is shaped as a cone and requires only surface area and maximum depth; "voldev" uses the volume development (Vd) parameter from Hakanson (1981) and Johansson et al. (2007). Vd is a dimensionless parameter that describes lake basin shape in relation to the volume of cone whose base area and height equal the surface area and maximum lake depth, it is estimated as Vd = Zmean/Zmax (Hakanson et al. 2000). Method "voldev’ requires lake surface area, mean and maximum depth.