图  1  10～30天带通滤波非绝热加热经验正交函数分解第一模态空间分布（实线表示3000 m地形）

Figure  1.  EOF1 (first mode of EOF) of 10–30-day filtered diabatic heating (the solid line represents the 3000-m terrain)

图  2  1979～2018年高原非绝热加热准双周振荡（QBWO）强度距平序列（柱状，对应左侧纵坐标）以及强度3～7年Butterworth带通滤波后的标准化时间序列（折线，对应右侧纵坐标）；折线上的大圆标记表示高、低指数年份

Figure  2.  Quasi-Biweekly Oscillation (QBWO) intensity anomaly of diabatic heating over TP for 1979–2018 (bars, corresponding to the left ordinate) and normalized time series of 3–7-year Butterworth band-pass filtered intensity (the solid line corresponds to the right ordinate); the large circle marks on the solid line indicate high and low index years

图  3  （a）高指数年和（b）低指数年对应的逐年PC1（10~30天非绝热加热EOF第一模态对应时间系数）标准化序列以及8个位相划分示意图（水平实线表示零线，水平虚线表示±0.75个标准差）

Figure  3.  Normalized PC1 (principal components corresponding to EOF1 of 10–30-day filtered diabatic heating) series of (a) high index year and (b) low index year and 8-phase division diagrams (the horizontal solid line represents the zero line; the horizontal dashed line represents ±0.75 standard deviations)

图  4  高指数年夏季10～30天降水合成（单位：mm d^–1）（打点区域表示通过90%置信水平检验区域）；红色矩形框表示长江以南关键区（23°～31°N，106°～120°E）

Figure  4.  Composite of 10–30-day rainfall for high index years (units: mm d⁻¹) (the dotted area is statistically significant at 90% confidence level ); the red rectangular box indicates the key area of the south of the Yangtze River (23°–31°N, 106°–120°E)

图  5  同图4，但为低指数年。上下两个红色矩形框分别表示江淮关键区（29°～34°N，112°～121°E）和华南关键区（23°～28°N，112°～121°E）

Figure  5.  As in Fig.4, but for the low index year. The upper and lower red rectangular boxes represent the key areas of Jianghuai (29°–34°N, 112°–121°E) and South China (23°–28°N, 112°–121°E), respectively

图  6  对RI_high指数回归的整层水汽通量（矢量，单位：kg m⁻¹ s⁻¹；黑色表示通过95%置信水平检验）和整层水汽通量散度（阴影，单位：10⁻⁵ kg m⁻² s⁻¹；仅绘制通过95%置信水平检验的区域）。其中字母“A”和“C”分别表示反气旋性、气旋性环流异常。d0即0天，表示与RI_high同期变化的环流异常；d−2即−2天，表示RI_high超前2天的环流异常

Figure  6.  Regression of vertically integrated water vapor flux (vector, units: kg m⁻¹ s⁻¹; the black vector is statistically significant at 95% confidence level) and the divergence of the water vapor flux (shaded, units: 10⁻⁵ kg m⁻² s⁻¹; only areas statistically significant at 95% confidence level are drawn) against RI_high. Letters “A” and “C” indicate anticyclonic and cyclonic circulation anomaly, respectively. D 0, day 0, means simultaneous regression with RI_high and circulation anomaly ; d−2, day −2, means 2-day lead from d 0

图  7  同图6，但为对RI_low指数的回归场

Figure  7.  As in Fig.6, but for the regression against RI_low

图  8  对RI_high指数回归的非绝热加热异常（阴影，单位：W m⁻²；仅绘制通过95%置信水平检验的区域）、500 hPa流函数（等值线间隔为1.0×10⁵ m² s⁻¹；虚线为负值并略去零线）和500 hPa水平风场（矢量，单位：m s⁻¹；黑色表示通过95%置信水平检验）。其中字母“A”和“C”分布表示反气旋性、气旋性环流异常

Figure  8.  Regression of diabatic heating anomaly (shading, units: W m⁻²; only areas statistically significant at 95% confidence level are drawn), 500 hPa flow function (isoline interval = 1.0 × 10⁵ m² s⁻¹, the dashed line corresponds to the negative value, and the zero line is omitted) and 500 hPa horizontal wind field (vector, units: m s⁻¹; the black vector is statistically significant at 95% confidence level) against RI_high. Letters “A” and “C” indicate anticyclonic and cyclonic circulation anomaly, respectively

图  9  同图8，但为200 hPa环流回归场（等值线间隔2.0×10⁵ m² s⁻¹；略去零线）

Figure  9.  As in Fig.8, but for 200 hPa (isoline interval=2.0 × 10⁵ m² s⁻¹; the zero line is omitted)

图  10  同图8，但为对RI_low指数回归场（等值线间隔为0.5×10⁵ m² s⁻¹；略去零线）

Figure  10.  As in Fig.8, but for the regression against RI_low (isoline interval = 0.5 × 10⁵ m² s⁻¹; the zero line is omitted)