Retrievals of Arctic sea ice melt pond depth and underlying ice thickness using optical data

Melt pond is a distinctive characteristic of the summer Arctic,which affects energy balance in the Arctic system.The Delta-Eddington model (BL) and Two-str Eam r Adiative transfer model (TEA) are employed to retrieving pond depth Hand underlying ice thickness Haccording to the ratio X of the melt-pond albedo in two bands.Results showed that whenλ=359 nm andλ=605 nm,the Pearson’s correlation coefficient r between X and His 0.99 for the BL model.The result of TEA model was similar to the BL model.The retrievals of Hfor the two models agreed well with field observations.For H,the highest r(0.99) was obtained whenλ=447 nm andλ=470 nm for the BL model,λ=447 nm andλ=451 nm for the TEA model.Furthermore,the BL model was more suitable for the retrieval of thick ice (0<H<3.5 m,R~2=0.632),while the TEA model is on the contrary (H<1 m,R~2=0.842).The present results provide a potential method for the remote sensing on melt pond and ice in the Arctic summer.