Assessing the Applicability of Random Forest, Stochastic Gradient Boosted Model, and Extreme Learning Machine Methods to the Quantitative Precipitation Estimation of the Radar Data: A Case Study to Gwangdeoksan Radar, South Korea, in 2018

Figure 6

Plots of rainfall rate estimations versus observations of tested models for precipitation event #4. (a) ZR1-L1 (RMSE: 2.88, R: 0.42). (b) ZR2-L1 (RMSE: 2.86, R: 0.43). (c) ZR3-L1 (RMSE: 2.88, R: 0.42). (d) ZR5-L1 (RMSE: 2.86, R: 0.43). (e) ZR1-L0 (RMSE: 3.05, R: 0.27). (f) ZR2-L0 (RMSE: 3.02, R: 0.3). (g) ZR3-L0 (RMSE: 3.05, R: 0.28). (h) ZR5-L0 (RMSE: 3.02, R: 0.3). (i) RF1 (RMSE: 2.93, R: 0.43). (j) RF2 (RMSE: 2.97, R: 0.42). (k) RF3 (RMSE: 2.89, R: 0.45). (l) RF5 (RMSE: 2.85, R: 0.46). (m) GBM1 (RMSE: 2.89, R: 0.42). (n) GBM2 (RMSE: 2.86, R: 0.42). (o) GBM3 (RMSE: 2.88, R: 0.42). (p) GBM5 (RMSE: 2.86, R: 0.43). (q) ELM1 (RMSE: 2.91, R: 0.39). (r) ELM2 (RMSE: 2.9, R: 0.41). (s) ELM3 (RMSE: 2.9, R: 0.4). (t) ELM5 (RMSE: 2.89, R: 0.41).

(a)

(b)

(c)

(d)

(e)

(f)

(g)

(h)

(i)

(j)

(k)

(l)

(m)

(n)

(o)

(p)

(q)

(r)

(s)

(t)