探討亞馬遜地區降水年際變化增加的因子

Abstract

亞馬遜地區擁有全世界面積最大的雨林，並且有豐富的生物多樣性，同時有地球之肺之稱，然而從觀測資料顯示，此處在1979到2014年降水年際變化增加，濕季變得更濕、乾季變得更乾，過去的研究探討造成這個現象的因子可能主要有自然變異和人為影響。自然變異之中以大西洋數十年週期振盪與太平洋十年振盪的影響最大，而人為影響中，砍伐森林和全球暖化都可能改變當地氣候，但是目前研究並沒有指出各個因子的貢獻程度為何，本研究利用全球氣候模式(National Center for Atmospheric Sciences, Community Earth System Model)設計四組理想實驗，分別為控制北大西洋和北太平洋的海溫，讓其海溫維持氣候平均值的季節循環，以及去除人為土地利用變遷與二氧化碳濃度升高的趨勢。 實驗結果指出在固定北大西洋海溫實驗中，降水季節性趨勢和控制組的差異最顯著，顯示北大西洋海溫影響亞馬遜降水的年際變化的重要性，我們同時做了一組移除北大西洋海溫上升趨勢的實驗，發現這組實驗對於降水年際變化的增加和固定北大西洋季節循環海溫實驗中有相同程度的改變，這結果指出北大西洋海溫的上升趨勢可能是造成亞馬遜地區的降水年際變化變大的主因。而在第六次耦合模式比對專案(Coupled Model Intercomparison Project Phase 6, CMIP6)中，當北大西洋暖化時，1979到2014年有14個模式同時顯示乾季變乾、濕季變濕的情形，然而在未來人為氣候變遷情境下，雖然北大西洋海溫有持續上升的趨勢，亞馬遜降水年際變化沒有增加的更多，顯示氣候暖化下的影響因子會隨著不同的氣候狀態而有所不同，未來研究將持續探討其背後不同的影響機制，對於未來亞馬遜雨林降水的變化以及其相對應的水、碳循環，可以有更準確的推估。

The Amazon precipitation annual range (AR) increases from 1979 to 2014. This change may result from natural variabilities and anthropogenic forcings (such as local land-use changes and global warming). The Atlantic Multidecadal Oscillation (AMO) and Pacific Decadal Oscillation (PDO) are two of the leading natural variabilities that significantly affect the long-term climate of Amazon. Amazon deforestation can cause local climate changes, including decreasing the mean precipitation. Furthermore, the increased atmospheric CO2 concentration can change patterns and characteristics of precipitation by radiative forcing and biophysical processes. However, the relative contributions from the natural and anthropogenic forcings on the Amazon precipitation AR remain unclear. This study uses National Center for Atmospheric Research (NCAR) Community Earth System Model (CESM) with a series of idealized experiments from 1970 to 2014 to explore each element's role in affecting Amazon AR by setting sea surface temperature (SST) as a climatological seasonal cycle over the North Atlantic and North Pacific, and fixing the land type and CO2 concentration at the level of 1970. The result shows that each factor contributes different magnitudes of changes in the precipitation trend in wet and dry seasons. In the North Atlantic climatological seasonal cycle of SST experiment, the precipitation AR trend over Amazon is the lowest. I further conduct the North Atlantic SST detrend experiment and find that North Atlantic SST warming trend seems to be the key to the increased AR over Amazon from 1979 to 2014, which can be revealed by the Coupled Model Intercomparison Project Phase 6 (CMIP6) model archive as well. In three-quarters (about 19 models) of the CMIP6 model archive under historical scenarios, the Amazon precipitation AR increases. In the future scenario, however, the larger North Atlantic SST warming does not necessarily lead to higher precipitation AR in the Amazon. Instead, the dry-season precipitation would keep decreasing, and the wet-season precipitation does not have a significant trend in the future, which deserves further research to better understand Amazon climate changes.