Research On The Streamflow Responses To Vegetation Cover Change And Climate Variability

This paper investigates the impacts of plantation expansion or clearing and climate variability on streamflow using two approaches:the sensitivity-based approach (including a non-parametric model and six Budyko framework based models) and the hydrological modelling approach (using Xinanjiang and SIMHYD models) for three medium sized catchments, Crawford River, Darlot Creek and Tinana Creek in Australia. The main focus and conclusions of this study are as follows:(1) Changing trend is one of the most important indexes which reflects changes of different elements in catchment hydrological recycle. The Mann-Kendall trend test, Spearman's trend test and linear regression trend test are used for testing changing trend of precipitation, areal potential evapotranspiration and streamflow time series, respectively. The effects of vegetation cover change and climate variability on streamflow are evaluated qualitatively. Then, the Flow Duration Curve (FDC) method is introduced to display frequency changes of streamflow between pre-plantation and post-plantation period. Comparison of FDCs in the annual streamflow from different periods indicates the important role of vegetation cover change.(2) It is a focus of study on effects of forest vegetation cover change on streamflow in modern forest hydrology research. The most basic and reliable approach for researching hydrological response to vegetation cover change is the catchment experiment. After analyzing changes of hydrological and meteorological elements, a framework for separating effects of vegetation cover change and climate variability on streamflow is developed using single catchment experiment. The sensitivity-based approach(including a non-parametric model and six Budyko framework based models) are used to estimate effects of climate change on streamflow. According to reliability and consistency analysis for each approaches, effects of vegetation cover change on streamflow are calculated correspondingly. There are several defects in the single catchment experiment such as streamflow measurement with long-period and high cost of managing the catchment. In order to overcome these problems, a reverse scenario is proposed in this paper to estimate effects of clearing on streamflow comparing to the plantation catchments experiments.(3) This paper presents a widely used statistical approach, the sensitivity based approach (using seven methods, one non-parametric method, six Budyko-framework based methods) to separating the effects of vegetation cover change and climate variability on streamflow. However, these statistical methods can only estimated the effects on annual time scale. For the sake of learning the effects of vegetation change and climate variability on streamflow process on daily, monthly time scale, two conceptual rainfall-runoff models, Xinanjiang model and SIMHYD model are used. Generalized Pattern Search (GPS) algorithm is introduced to calibrated models. The hydrological modelling is undertaken in two ways. Firstly, the two hydrological models are calibrated using streamflow data from the pre-plantation period and the calibrated parameters are used to simulate runoff for the post-plantation period. The difference between the observed and model simulated runoff for the post-plantation period is used to quantify the impacts of the increase in plantations and climate variability between the two periods. Secondly, the two hydrological models are calibrated using streamflow data from the post-plantation period and the calibrated parameters are used to simulate runoff for the pre-plantation period. The difference between the observed and model simulated runoff for the pre-plantation period is used to quantify the impacts of the reduction in plantations (considered equivalent to plantation'clearing'in this study) and climate variability between the two periods.(4) Hydrological cycle can be significantly influenced by vegetation cover change through affecting soil water storage, canopy interception and evapotranspiration. However, vegetation processes are seldom considered in traditional lumped conceptual rainfall-runoff models. This paper incorporates the remotely sensed AVHRR-LAI data into Xinanjiang model and SIMHYD model by way of revising the original three-layer and one-layer evapotranspiration submodel using Penman-Monteith equation, respectively. The so called Xinanjiang-ET model and SIMHYD-ET model are calibrated and validated using streamflow data from Crawford River catchment. The results indicate that the calibrations as well as validations for the revised versions of Xinanjiang model and SIMHYD model can be improved by considering remotely sensed AVHRR-LAI data.