Impact of river runoff into the ocean on climate in a coupled model

Abstract

In order to study the impact of global runoff, sensitivity experiment (NoRiv) was carried out by shutting down the entire runoff into the ocean for a period of 200 years. The changes in NoRiv, compared to the reference simulation (CR), shows the impact of global runoff (NoRiv minus CR). The evolution of mean ocean SSS in the major ocean basins show that the upper– ocean reached a quasi-stabilized stage in the first 100 years in the NoRiv. Globally, when there is no runoff, the surface ocean turns saltier with the largest increase occurring off the mouths of major river systems. High salinity patches of more than 2 psu are found near the mouths of Amazon, Congo, Ganga–Brahmaputra, Yangtze, etc. The upper–ocean currents distribute the positive salinity anomalies generated near the river mouth of NoRiv to far away regions and open ocean. In accordance with our results using CCSM3.0 model, CESM1.0 also shows that the impact of runoff into the ocean on SSS is significantly higher in the extratropical ocean than in the tropics, though the tropical runoff is considerably higher (Vinayachandran, Jahfer, and Nanjundiah, 2015). Arctic Ocean recorded the highest rise in salinity as the ratio of runoff to the surface area of the basin is highest. The open ocean regions of Pacific exhibited the least change in salinity due to the vastness of the basin. We further analyzed the significance of this rise in SSS on the upper–ocean temperature, air–sea interaction, and rainfall. A saltier ocean in the NoRiv deepens the mixed layer of the upper–ocean owing to the increase in surface density. Though the mean MLD in the NoRiv is found to be deeper than the CR, the SST response to this change in ML is complex. In the simulation without runoff, the northern hemisphere oceans recorded a comparatively higher rise in surface temperature, mostly in the Atlantic Ocean. The change in SST in the NoRiv is primarily due to the weakening of upper–ocean stratification or due to changes in air–sea interaction in the coupled system. The equatorial Ocean of the Pacific Ocean recorded a mean cooling (of 0.2oC). But, the magnitude of the impact is relatively weaker in CESM1.0 as compared to CCSM3.0. The evolution of SST anomaly in the equatorial Pacific is a consequence of complex air–sea interaction. The appearance of cooler SST anomalies in the eastern equatorial Pacific of NoRiv generates a zonal pressure gradient resulting in an enhancement in easterlies. This, in turn, enhances wind– driven upwelling in the eastern equatorial Pacific leading to the formation of a La Niña–like SST anomaly in the equatorial Pacific. This La Niña–like SST anomaly in the equatorial Pacific has the potential to alter the Indian summer monsoon, one of the most dominant and regular features in the present–day tropical climate. It is well established that the cooler phase of equatorial Pacific, the La Niña, favors a higher rainfall over the Indian subcontinent. Existing theories on the teleconnection between the Indian monsoon and ENSO states that the summer– time changes in the equatorial Pacific SST results in convective anomalies in the Pacific Ocean that can impact the rainfall over the Indian subcontinent (Shaman and Tziperman, 2007). Chapter 6. Summary and Conclusions 151 A cooler Pacific, La Niña phase, favors negative convection anomalies in the tropical Pacific which leads to atmospheric convergence over Indian landmass and thereby affects the westward propagating jet termed as the North African–Asian jet (NAA; Shaman and Tziperman, 2007). The resultant negative vorticity anomalies over the Asian landmass during a La Niña phase lead to overall warming over the region that reinforces the meridional temperature gradient between equatorial Indian Ocean and landmass over Asia (Shaman and Tziperman, 2007). However, we find that the response of SSS, SST, and rainfall to the global runoff in the recent version (CESM1.0) is weaker than the earlier one (CCSM3.0). A higher ocean–land temperature gradient in the NoRiv is found to be favorable for increased intraseasonal activity over the Indian region leading to a stronger summer monsoon (Jiang, Li, andWang, 2004). Thus the improved monsoon in the absence of global runoff is thus a result of cooling in the equatorial Pacific and an upper–level vorticity guided strengthening of the meridional temperature gradient.

The 0AMZ experiment is similar to the global runoff experiment (NoRiv) except that the runoff into the ocean from the Amazon river is shut down. The last 100 years of the model simulation shows that the Amazon runoff is a vital component of the large–scale, low–frequency oscillations in the Atlantic Ocean. When there is no input from the Amazon river, saltier water from the river mouth disseminates over the North Atlantic Ocean as far as the deepwater formation sites in the extratropics. Consequently, the upper–ocean mixed layer over the deep–convection zones in the Labrador and GIN Sea region deepen. AdeeperMLimplies strengthening of deepwater formation rate and a reinforced Atlantic meridional overturning circulation (AMOC). On a longer timescale, the atmospheric response to Amazon runoff is found to be consistent with the well established global impacts during a positive phase of AMOC. Chapter 6. Summary and Conclusions 152 A stronger AMOC brings cooler subsurface water to the surface in the southern tropical Atlantic Ocean by enhancing the upwelling rate. This cooler upper–ocean water is carried to the north of the equator by the cross–equatorial currents. Consequently, the upwelling branch of Hadley Cell in the equatorial Atlantic Ocean becomes weaker. This weakening of Hadley Cell has a negative feedback on the atmospheric meridional cells in the subtropics and extratropics. The North Atlantic Oscillation, the strength of which depends on the strength of winter–time (DJF) meridional circulation in the extratropical Atlantic Ocean, thus shifts to its negative phase. Though the impact of Amazon runoff on rainfall during winter is relatively weaker in the tropics, the landmass adjacent to the extratropical Atlantic Ocean shows significant responses. We found that the runoff–induced changes in surface temperature, winds, and rainfall over the United States and Europe are similar to the anomalies during a negative phase of NAO. The winter storms over northern Europe turn weaker along with a reduction in local rainfall. However, southern Europe and eastern Canada received higher rainfall in the absence of Amazon runoff, whereas the rainfall over the eastern US decreased. In the tropical Atlantic Ocean sector, the summer–time rainfall (JJA) response exhibits a see–saw pattern with a higher rainfall over the northern half and lower rainfall to the south. This spatial pattern of rainfall anomaly is a typical feature of feedback during a positive phase of AMOC. During a stronger AMOC, the warm surface waters of the southern tropical Atlantic is pushed to the northern tropics, and the cooler waters dominate over the south and along the equator. The SST anomalies thus formed is strikingly similar to the surface ocean signature during a positive Atlantic Multidecadal Variability (AMV), which is believed to be a subset of AMOC. The changes in rainfall in the 0AMZ is closely tied to the spatial pattern of SST anomaly such that a warmer northern equatorial Atlantic Ocean favors a higher rainfall over the region and vice versa. Thus when there is no input from the Amazon river, the AMOC and AMV shifts to a positive phase and thereby affect the surface temperature and rainfall over the tropical Atlantic Ocean during boreal summer.