全球SPEI数据的应用及其与环流因子的多尺度分析 Application of Global SPEI Database and Its Multi-Scale Correlation with Circulation Factors

以石羊河上游水源区为例，运用交叉小波对标准化降水蒸散指数(SPEI)数据集数据、气象资料、太阳黑子数以及多条环流因子资料进行多尺度分析。结果表明：研究区上游SPEI值与临近气象站气象资料在多时间尺度上具有稳定的显著小波相关性，验证了全球SPEI数据集在该区的适用性。较短时间尺度SPEI以2~3年尺度的振荡周期为主，较长时间尺度SPEI表现出较强的8~11年尺度平稳、显著的振荡周期。此外，SPEI与太阳黑子数和环流因子存在多时间尺度的显著关联性：各时间尺度SPEI与太阳黑子数在整个研究期内具有极显著的8~12年主共振周期；较长时间尺度SPEI与NAO和PDO在准10年共振周期上显著相关(1970s~1990s)；而ENSO (El Nino-Southern Oscillation)对SPEI的影响主要表现在短时间尺度的高频段上，存在准3年(1960s)和4~6年(1985~1995年)显著性的共振周期。 Taking upper reaches of Shiyang River as an example, the multi-time scale cross-wavelet transformation was used to study SPEI, sunspot, climate factors (precipitation, maximum temperature and hours of sunshine) and large-scale circulation factors (PDO, NAO, Nino3.4 and SOI) as well as their relationships in the upper reaches of the Shiyang River. The results show that there is a significant correlation between SPEI and climate factors utilizing the method of cross wavelet coherence, and the precipitation change plays a major role in SPEI change. The applicability of global SPEI database in this region was tested. Results of continuous wavelet transformation show that SPEI12 has significant 2 - 3 years periods; SPEI48 has significant 8 - 11 years periods. The significant coherence was found between SPEI at different time-scales and sunspot and four large-scale circulation factors. There are the common patterns of 8 - 12 years oscillation circle between SPEI at different time-scales and sunspot during the whole period (1950-2000). NAO, PDO and SPEI48 in the upper reaches are significant correlated with at the scales of quasi-10 years during the period from 1970 to 1999. ENSO (El Nino-Southern Oscillation), SOI and SPEI1 (1 month scale SPEI) or SPEI12 (12 months scale SPEI) in the upper reaches are significant correlated with at the scales of quasi-3 years (1960’s) and 4 - 6 years during the period from 1985 to 1995.