Hydrological Characteristics And Evolutions Of Tropical Mountain Rainforest Ecosystem In Jianfengling, Hainan Island

Tropical mountain rainforest in Jianfengling is one of the most intact tropical rainforest ecosystems in China. It plays an important role in adjusting regional water recycle and water resources, keeping ecological balance of terrestrial ecosystems. In the end of 20th century, Zhou Guangyi and Cheng Bufeng had carried out a series of forest hydrological researches in Jianfengling. However, long term hydrological rule of tropical mountain rainforest ecosystem is still unknown. At present the impacts of global climate on water resources become a focus of our society. It is necessary to discuss long term hydrological evolutions of forest. This paper analyzed characteristics of rain and runoff, evolutions of tropical mountain rainforest hydrology, and impacts of climate change on runoff, based on the hydrological data(1989-2007)and climatic data(1957-2007)of Jianfengling national forest ecosystem research station. It helps in understanding the hydrological function of tropical rainforest, and the impacts of climate change on runoff.(1)Rainfall CharacteristicsAnnual rainfall of Tropical mountain rainforest changed from 1 305.5 mm to 3 662.3 mm, with the mean of 2 478.9 mm. May to October was the rain season in Jianfengling, and its rainfall accounted for 87.74% of annual rainfall. November to April of the next year was the dry season, and its rainfall accounted for only 12.26% of annual rainfall. August was the most abundant rainfall month, and the next was July, and then September. Rainfall of this three month accounted for 20.1%, 19.5% and 16.8% of annual rainfall, respectively. One significant characteristic of rainfall in Jianfengling was the rainstorm. Rainfall of huge rainstorm accounted for 31.3% of annual rainfall, and rainfall of rainstorm accounted for 17.9% of annual rainfall. Usually, rainstorms occurred in July to September, among which August was the most, there were 2.7 rainstorms per year in this month.(2)Runoff Characteristics Runoff of tropical mountain rainforest in Jianfengling was calculated and separated by digital filters. It proved that digital filters method was very useful. Annual mean base flow and quick flow were 673.6 mm and 721.3 mm, and mean baseflow index was 0.48. Annual mean runoff was 1 394.9 mm, among which rain season was 1 103.3 mm, and dry season was 291.6 mm. Runoff of April was the lowest in a year, with the value of 17.7 mm, and runoff of September was the highest, with the value of 349.4 mm. There were 6 abundant runoff years, 8 insufficient runoff years in 1989-2007, and the abundant runoff year and insufficient runoff year occurred almost alternately.(3)Water balance characteristicsCanopy interception was an important factor in water balance of tropical mountain rainforest in Jianfengling. Annual mean throughfall, stem flow and canopy interception were 2 296.3 mm, 205.5 mm and 409.2 mm, respectively. Annual mean throughfall rate, stem flow rate and interception rate were 78.7%, 7.1% and 14.2%, respectively. Surface flow was tiny, and annual mean surface runoff was 17.1 mm, which accounted for only 0.6% of annual rainfall. For tropical mountain rainforest ecosystem in Jianfengling, annual mean rainfall input was 2 453.6 mm, and output runoff was 1 245.4 mm, and evapotranpiration was 11.8.2 mm. Therefore, annual mean runoff coefficient was 0.53, and evapotranspiration coefficient was 0.47.(4)Evolutions of hydrological time seriesThere was not significant trend in annual rainfall time series and runoff time series, but there were increasing or decreasing trends in some stage. Jump points of annual rainfall time series and runoff time series were not significant. Wavelet analysis showed the obvious periodic variation appeared in time scale of 20 years, 10 years and 4 years for annual rainfall time series, and 10 years and 4 years for annual runoff time series. scale of 10 years may relate to the periodic variation of sun, and scale of 4 years may relate to ENSO. A 2 steps aotoregressive model was established to simulate annual runoff time series.(5)Climate change and its impacts on runoff There was a warming trend in annual air temperature time series of Jinagfengling, and wavelet analysis showed the obvious periodic variation appeared in time scale of 10 to 12 years, 5 to 6 years, and 2 to 4 years. 1960s-1970s was a cold period. Enter in 1980s, annual air temperature increased continually, with the increasing rate of 0.15℃/10 a. Enter in 21th centuries, annual temperature increased more significantly, and annual mean air temperature in 2001-2006 was 0.4℃higher than that of 1990s. 1979 was a jump point of the time series, and the mean annual air temperature in 1957-1978 was 24.4℃, but that of 1979-2007 was 25.0℃.There was a decreasing trend and a periodic variation in time scale of 10 years and 5 to 6 years in annual evaporation time series. The decreasing trend of Shiyanzhan was abrupt, but that of tropical mountain rainforest was slow. It may be own to the good protection of forest. The decreasing trend of evaporation was the integration of the decreasing trend of wind speed and sunshine hours, and the increasing trend of air temperature.Among all climate factors that affected runoff, rainfall affected runoff most greatly. The impact of air temperature was tiny compared to rainfall. However, the warming of air temperature had increased extreme weather events. And extreme weather such as typhoon had great impacts on rainfall and runoff. Simulation of impacts of annual temperature on runoff showed that when annual temperature increased 0.1℃, annual runoff decreased 6.3 mm to 6.7 mm. A CAR model was established in order to simulate impacts of multivariable on runoff, and it was better in forecasting the impacts of rainfall, temperature and evaporation on runoff.Innovative points:(1)Tropical mountain rainforest runoff was separated by digital filters, which was an automatic and repeatable method. Annual variations and monthly variations of base flow and quick flow were analyzed.(2)Trend, abrupt change, and periodicity of hydrological and climatic time series were analyzed by wavelet analysis and other time series methods. And Characteristics of these time series were clarified. (3)A CAR model was used to simulated the impacts of multivariable climatic factors. And this model was useful to analyze and forecast the impacts of climate change on runoff, based on the change characteristics of runoff and climatic time series.