The impact of rainfall on the sea surface salinity: a mesocosm study

Sea surface salinity may serve as a tracer for freshwater fluxes because it is linked to evaporation and precipitation that force the freshwater balance of the ocean’s surface. The relationship between freshwater fluxes and salinity anomalies in the upper few centimeters remains widely unknown. In a mechanistic approach, we investigated how these anomalies develop by conducting experiments with artificial rain over a large basin. We measured conductivity and temperature at different depths and rain characteristics (intensity, rain temperature, droplet sizes, and velocities). In the absence of turbulence, the rain causes a strong salinity change of up to 6.02 g kg\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$^{-1}$$\end{document}-1 in 0–4 cm depth. At the highest rain intensity of 56 mm h\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$^{-1}$$\end{document}-1, salinity changed thrice as fast as at an intensity of 18 mm h\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$^{-1}$$\end{document}-1. At the sea surface microlayer (first millimeter of the surface) the anomalies are always highest and reached a maximum of 14.18 g kg\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$^{-1}$$\end{document}-1 . With mechanical mixing, salinity changes were less pronounced (maximum SML salinity anomaly: 6.17 g kg\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$^{-1}$$\end{document}-1), and freshwater was mixed fast with the existing seawater body. In general, our study shows that freshwater remains in the upper few centimeters, and even with induced turbulence, are not mixed below 16 cm.

The rainfall amount was measured with the ow meter during the experiment.Sbefore and Safter represent the mean absolute salinity before and after the rainfall phase during the 15-minute periods, respectively.V 0−2cm is the volume of the water mass inside the tank of the rst 2 cm.

Calculation of fractional change of salt content
We used the following equation to calculate the fractional change of salt content at the dierent depths after the rainfall of the experiment with a turbulent-free waterbody: The initial salinity is described with S 0 and the salinity after the artial rain with S R .The salinity of the artical rain water (S rain ) is expexted to be 0. A result of 1 indicates no change in salt content.During the phase of no mixing, stratication developed after rainfall at all three intensities of 18, 28, and 56 mm h −1 .The positive ∆T at the surface shows that the temperature of the articial rain was higher than that of seawater.The anomalies of N2 with 28 mm h −1 are higher compared to the anomalies of N3 with 56 mm h −1 .The mean temperature of the rain was dierent with 17.54 °C, 17.44 °C, and 15.27 °C at the intensities of 18 mm h −1 , 28 mm h −1 , and 56 mm h −1 , respectively.

Figure S1 :
Figure S1: Distribution of droplet sizes and velocities of the three nozzle types N1, N2, and N3 and rain intensities 18 mm h −1 , 28 mm h −1 , and 56 mm h −1 directly under the nozzle.The colors show the mean number of droplets of the size and velocity categories of 1 minute.

Figure S2 :
Figure S2: Time series of temperature changes during dierent rain scenarios with the rst and second runs of experiment one (turbulence-free water body).The light blue rectangle indicates the 15-minute rainfall period.

Figure S3 :
Figure S3: Time series of salinity and temperature changes during dierent rains in experiment two (turbulence-mixed water body).The light blue rectangle indicates the 15-minute rainfall period.The dotted lines indicate the start and stop of the 45 minutes with a certain pump level of 0 %, 0.5 %, and 1 %.

Figure S4 :
Figure S4: Contour plot of ∆T at dierent depths during three rainfall scenarios in the second experiment with a turbulence-free and mixed water body.Shown are the lowest intensity of 18 mm h −1 and N1 (a-b), the medium intensity of 28 mm h −1 and N2 (c-d), and the highest intensity of 56 mm h −1 and N3 (e-f) with a pump level of 0 % and 1 %.The start and end times of the precipitation phase are indicated with a dashed line (minutes 1630).The black isolines show the progression of equal temperatures with time and depth.

Figure S5 :
Figure S5: Contour plot of ∆D at dierent depths during three rainfall scenarios of the second experiment with a turbulence-mixed water body.Shown are the lowest intensity of 18 mm h −1 and N1 (a-b), the medium intensity of 28 mm h −1 and N2 (c-d), and the highest intensity of 56 mm h −1 and N3 (e-f) with pump levels of 0 %, 0.5 %, and 1 %.The start and end times of the precipitation phase are indicated with a white dashed line (minutes 1630).The black isolines show the progression of equal sigma-t densities with time and depth.

Figure S6 :
Figure S6: Scatter plot showing the correlation between the rainfall intensity and the maximum salinity (a) and temperature anomalies (b) at a depth between 0 and 2 cm during the rst experiment and the turbulent-free parts of the second experiment.n = 13.

Figure
Figure S7: 15 Positions of the optical rain sensor 1A3E.Each red rectangle represents the area of the laser generated by the distrometer to measure rainfall intensity and droplet properties, such as drop sizes and velocities.The red star indicates the position of the nozzle above the calibration area.

Figure S8 :
Figure S8: Contour plot of ∆S and ∆T at dierent depths of the second experiment with a turbulence-mixed water body.Shown are ∆S (a) and ∆T (b) of the lowest intensity of 18 mm h −1 and N1, the medium intensity of 28 mm h −1 and N2 with ∆S (c) and ∆T (d), and the highest intensity of 56 mm h −1 and N3 with ∆S (e) and ∆T (f) with a mean TKE of 3.37 x 10 −4 m −2 s −2 (pump level: 0.5 %).The start and end times of the precipitation phase are indicated with a white dashed line (minutes 1630).The black isolines show the progression of equal sigma-t densities with time and depth.

Figure S9 :
FigureS9: Timeseries of a test run with captured salinity and temperature by to dierent mounted CTDs at 8 cm depth with an rain intensity of 56 mm h −1 with N3.One is mounted horizontal and one vertical in the water of the mesocosm tank.Rain fall was applied for 30 minutes to the tank.