CAJ | 학술논문

降雨侵蚀力反映了降雨对土壤侵蚀的潜在能力，研究其时空变化特征对流域土壤侵蚀监测、评估、预报和治理等工作具有重要意义。根据珠江流域43个气象站1960－2012年逐日降雨资料计算各站点降雨侵蚀力，采用线性回归， Mann-Kendall方法，小波分析和Kriging插值等方法对流域降雨侵蚀力进行了时空变化分析。结果表明：珠江流域多年平均降雨侵蚀力值的分布范围为1858.0～14656.6 MJ·mm/(hm2·h)，平均值为7177.1 MJ·mm/(hm2·h)，与多年平均降雨量极显著相关（相关系数0.952，P<0.01），空间分布规律与多年平均降雨基本一致，即总体上均呈从东到西逐渐递减的规律，被统计站点的降雨侵蚀力随着经度增加而增加，但随纬度增加而减少；流域年、季节、汛期和非汛期降雨侵蚀力变化趋势均不显著，均没有发生显著的突变，其中春、秋两季降雨侵蚀力呈下降趋势，其余时间段呈上升趋势；珠江流域大部分地区年降雨侵蚀力呈上升的趋势，其中韶关站点上升显著，沾益站、风山站、河池站、百色站、柳州站、融安站和桂林站的冬季降雨侵蚀力同样上升显著，这些地区面临的水土保持压力较大；流域年均降雨侵蚀力变化主周期为3.8 a，且存在2.0～7.0 a的振荡周期。研究结果可为珠江流域的水土保持、农业和生态保护，灾害控制等工作提供科学决策依据。
강우침식력반영료강우대토양침식적잠재능력，연구기시공변화특정대류역토양침식감측、평고、예보화치리등공작구유중요의의。근거주강류역43개기상참1960－2012년축일강우자료계산각참점강우침식력，채용선성회귀， Mann-Kendall방법，소파분석화Kriging삽치등방법대류역강우침식력진행료시공변화분석。결과표명：주강류역다년평균강우침식력치적분포범위위1858.0～14656.6 MJ·mm/(hm2·h)，평균치위7177.1 MJ·mm/(hm2·h)，여다년평균강우량겁현저상관（상관계수0.952，P<0.01），공간분포규률여다년평균강우기본일치，즉총체상균정종동도서축점체감적규률，피통계참점적강우침식력수착경도증가이증가，단수위도증가이감소；류역년、계절、신기화비신기강우침식력변화추세균불현저，균몰유발생현저적돌변，기중춘、추량계강우침식력정하강추세，기여시간단정상승추세；주강류역대부분지구년강우침식력정상승적추세，기중소관참점상승현저，첨익참、풍산참、하지참、백색참、류주참、융안참화계림참적동계강우침식력동양상승현저，저사지구면림적수토보지압력교대；류역년균강우침식력변화주주기위3.8 a，차존재2.0～7.0 a적진탕주기。연구결과가위주강류역적수토보지、농업화생태보호，재해공제등공작제공과학결책의거。
Soil erosion is recognized as one of the most serious, global ecological environmental crises in progress today. A rainfall-runoff erosivity factor, combined with the effects of duration, magnitude and intensity of rainfall event, can be used to measure the rain’s potential ability to cause erosion. In this paper, the rainfall erosivity model proposed by the Chinese scholar Zhang Wenbo was used to calculate the rainfall erosivity. Taking the Pearl River basin as the study case, daily rainfall data from 1960 to 2012 in 43 meteorological stations were applied in the model. Methods of linear regression, Mann-Kendall, wavelet analysis and Kriging interpolation were applied to analyze the spatial and temporal variations of rainfall erosivity. The results showed that the range of annual rainfall erosivity in the Pearl River basin was 1858.0-14656.6 MJ·mm/(hm2·h) with an average value of 7177.1 MJ·mm/(hm2·h). The average annual rainfall erosivity decreased from east to west in general. Larger values mainly appeared in most areas of Pearl River Delta region, Dongjiang River basin and Beijiang River basin, but the values in Nanpanjiang and Beipanjiang River basin which are the upstream regions of the Pearl River basin were smaller. The distribution of average annual rainfall erosivity was similar with the average annual rainfall and there was a strong correlation (R=0.95, P<0.01) between them. Moreover, the average annual rainfall erosivity generally increased with the increasing of longitude (R=0.712, P<0.01), but decreased with the increasing of latitude (R=0.449, P<0.01). Trends of rainfall erosivity were not significant among years, four seasons, flood and non-flood seasons and no significant mutations occurred in these periods. Among them, the rainfall erosivity showed a slight downward trend in spring and autumn, but a slight upward trend in other periods. Among the periods of upward trend, the rainfall erosivity rising in summer was the fastest with a climbing speed of 11.251 MJ·mm/(hm2·h·a) and the average summer rainfall erosivity reached up to 5 414.530 MJ?mm/(hm2?h). The rainfall erosivity rising in year was the second fastest with a climbing speed of 8.469 MJ?mm/(hm2?h) and the average annual rainfall erosivity reached up to 10235.962 MJ·mm/(hm2·h). In winter, the rainfall erosivity of 39 meteorological stations accounting for about 90.7%of the total stations showed an upward trend, suggesting that the erosivity in winter rose overall. The mutation analysis of rainfall erosivity in the Pearl River basin indicated that the average annual rainfall erosivity sequences of the 7 sub-regions and the whole basin did not have significant mutations. The annual rainfall erosivity of most areas in the Pearl River basin showed significant upward trends (P<0.05), especially at Shaoguan station. In winter, the annual rainfall erosivity at Zhanyi, Fengshan, Hechi, Baise, Liuzhou, Rong’an and Guilin stations also showed significant upward trends. Evidently, the regions represented by these stations faced great pressure in the water and soil conservation. The rainfall erosivity sequence from 1960 to 2012 had the periods of 2.3, 3.8, 6.9, 12.7 and 23.4 a. Among them, only 2.3 and 3.8 a passed the red noise test at confidence level of 95%. The peak value of wavelet variance in 3.8 a was larger than 2.3 a, which suggested that 3.8 a was the main period. The red noise test also indicated that there was an oscillation period of 2.0-7.0 a in the basin. Generally speaking, the rainfall erosivity in the Pearl River basin showed an exacerbated trend, and therefore water and soil conservation should be well prepared. This study has the potential to provide an important reference for soil and water conservation planning, agricultural protection, ecological protection and disaster control in the Pearl River basin.