1. Introduction
The role of the Andes in supplying water through orographic precipitation is of vital importance for adjacent lowlands in Chile and western Argentina. In winter, the subtropical central Andes (SCA; 30°–37°S) is mostly affected by the northern flank of storm tracks (Trenberth 1991; Hoskins and Hodges 2005), when cyclones moving eastward produce strong cross-barrier flow that results in upslope precipitation on the windward slope and rain shadow effect on leeward slopes. These orographic effects accentuate in mountain ranges oriented perpendicular to the prevalent horizontal flow and, hence, have great influence on the climate in adjacent areas, producing strong windward–leeward gradients in vegetation and water availability [e.g., New Zealand Alps (Griffiths and McSaveney 1983; Wratt et al. 2000), the Cascades in Oregon (Smith et al. 2005), and the southern Andes (Smith and Evans 2007)].
Orographic effects on the horizontal flow result in a very complex distribution of precipitation across the mountain range. A dense observation network is necessary to depict the spatial precipitation pattern, which is a real limitation in the South American Andes. For example, Falvey and Garreaud (2007, hereafter FG07) found an orographic precipitation enhancement close to 2–3 in SCA using mostly river discharge estimates. Between 40° and 48°S in the southern Andes, Smith and Evans (2007) reported the highest drying ratio1 values ever found in a mountain range using stable isotope data from stream water. Over other more-sampled mountain ranges, the maximum precipitation was identified on the windward slope of the high Cascade or Sierra Nevada ranges (Colle and Mass 2000; Smith et al. 2005; Leung and Qian 2003) and/or over the crest of the low Oregon coastal mountains (Colle and Mass 2000) or low New Zealand Alps (Sinclair et al. 1997). Using one of the densest rain gauge networks on a mountain range, Frei and Schär (1998) identified finescale spatial variability as the most prominent characteristic in precipitation fields over the European Alps, with precipitation enhancement on the upslope peaks and shielding in inner valleys. At the small ridge–valley scale, strong precipitation gradients were also documented over the Olympic Peninsula in Washington; these remained relatively constant on time scales ranging from annual to single event, thus suggesting the dominant role of the topography in determining the spatial precipitation pattern (Anders et al. 2007; Minder et al. 2008). By contrast, the Andes are a data-poor region. Therefore, we focus on variations in the cross-mountain direction of precipitation over the (still poorly understood) broad scale of ~50 km, distinguishing between robust cross-barrier zones of the low-windward side, windward slope, immediate leeward slope, and low-leeward side.
The orographic precipitation pattern depends on synoptic forcing, including air mass stability, moisture content, and direction and strength of wind, which in turn interacts with the topography. Junker et al. (2008) and Pandey et al. (1999) showed that deeper cyclones located off the western U.S. coast lead to stronger winds and moisture fluxes against the Sierra Nevada, resulting in heavier precipitation. Based on data collected along and off the California coast during the California Landfalling Jets Experiment (CALJET) and the Pacific Landfalling Jets Experiment (PACJET) (Ralph et al. 1999), Ralph et al. (2004, 2005a) documented that water vapor transport concentrates over an extensive and narrow region of high water vapor content associated with the low-level jet in the broader warm and pre-frontal zone of the polar front (i.e., the “warm conveyor belt”; Browning 1990). This long and narrow corridor of water vapor above the ocean accounts for essentially the total meridional transport at middle latitudes, so it has been named “atmospheric river” (Zhu and Newell 1998). Recent composite studies of atmospheric rivers have demonstrated their crucial role in modulating heavy orographic rainfall, snowpack variability, and flooding in western North America (Ralph et al. 2004, 2005a, 2006; Neiman et al. 2008). Given the significant contribution of atmospheric rivers to the extreme precipitation in the western United States, the results of the Hydrometeorology Testbed (HMT) project-West field programs, conducted by the National Oceanic and Atmospheric Administration (NOAA) since 2005 (Ralph et al. 2005b, 2010), may be applicable to other north–south mountain ranges such as the Andes.
Because of sparse distribution of surface stations in the Andes of South America, studies of winter orographic precipitation are limited. Recent modeling case studies have documented that local airflow characteristics and precipitation patterns result from the interaction between the synoptic-scale flow and the topography of SCA (Barrett et al. 2009; Viale and Norte 2009, hereafter VN09). A climatological approach of winter precipitation and their associated synoptic conditions have been addressed by FG07, but is limited to the low-windward side and windward slope of the subtropical Andes. Our study extends the climatological approach of winter orographic precipitation to the leeward slope and low-lee side of SCA, making use of the less sparse network of precipitation gauges available over the mountains and both adjacent low sides between 30° and 37°S. We also explore the synoptic and regional air mass features up- and downstream of the barrier that accompanied heavy orographic precipitation events, including the possible linkage with landfalling atmospheric rivers on the western coast of South America.
The remainder of this article is structured as follows: the data and topographic features, as well as the precipitation-event dataset and their composite methodology, are described in the next section. In section 3 we examine the spatial, seasonal, and daily distribution of winter precipitation over the mountain range and their surroundings. The synoptic and regional conditions during the heavy precipitation events and their links with atmospheric rivers are analyzed in section 4. Our main results are discussed and summarized in section 5.
2. Data and methods
a. Surface data and geographic setting
In the last 30 years, the number of the national weather services stations in Argentina [Servicio Meterológico Nacional (SMN)] and Chile [Dirección Meteorologica de Chile (DMC)] decreased considerably, including in the Andean region. Because of this, our analysis is based on the winter precipitation database (April–September) for the period 1970–76 since it provides the densest rain gauge network available both in the mountains and in adjacent lowlands. The stations used in this study are shown in Fig. 1. Data from at least 12 mountain stations were used—4 with daily and 8 with monthly records (Table 1). Mountain stations (i.e., at elevations > 2000 m) were mostly located in river valleys and canyons in the Andes—where in situ observers registered snow depth and density—to calculate snow water equivalent. A total 80 stations were used, of which 51 stations are located west (Chile) and 29 east (Argentina) of the Andean divide. The 51 Chilean stations comprise 2 daily and 49 monthly precipitation series. The series were provided by the DMC, the National Environment Center (CENMA), and the General Water Direction (DGA), though most of them come from the Global Historical Climatology Network (GHCN). The 27 Argentinean stations consist of 8 monthly observation stations belonging to the General Irrigation Office (DGI) and 19 daily stations from the SMN.
Orographic region investigated in this study (elevations in m; above 500 m are shaded) and the stations used (filled and empty circles represent stations with daily and monthly precipitation data, respectively). The stations enclosed by the vertical rectangle with dashed white lines correspond to high-mountain locations (i.e., alt > 2000 m), while horizontal rectangles with continuous lines correspond to stations used for cross-barrier plots in Fig. 4. The rawinsonde location of QUI station is indicated by a black diamond on the Pacific coast. The low-lying lee stations used in Table 6 display their ID name.
Citation: Journal of Hydrometeorology 12, 4; 10.1175/2010JHM1284.1
Orographic region investigated in this study (elevations in m; above 500 m are shaded) and the stations used (filled and empty circles represent stations with daily and monthly precipitation data, respectively). The stations enclosed by the vertical rectangle with dashed white lines correspond to high-mountain locations (i.e., alt > 2000 m), while horizontal rectangles with continuous lines correspond to stations used for cross-barrier plots in Fig. 4. The rawinsonde location of QUI station is indicated by a black diamond on the Pacific coast. The low-lying lee stations used in Table 6 display their ID name.
Citation: Journal of Hydrometeorology 12, 4; 10.1175/2010JHM1284.1
Orographic region investigated in this study (elevations in m; above 500 m are shaded) and the stations used (filled and empty circles represent stations with daily and monthly precipitation data, respectively). The stations enclosed by the vertical rectangle with dashed white lines correspond to high-mountain locations (i.e., alt > 2000 m), while horizontal rectangles with continuous lines correspond to stations used for cross-barrier plots in Fig. 4. The rawinsonde location of QUI station is indicated by a black diamond on the Pacific coast. The low-lying lee stations used in Table 6 display their ID name.
Citation: Journal of Hydrometeorology 12, 4; 10.1175/2010JHM1284.1
Coordinates and heights of the stations with daily data in low-windward and -leeward sides; all stations located on high mountain (i.e., alt > 2000 m; bold here and enclosed by white rectangles in Fig. 1) used in the study for wintertime (Apr–Sep) 1970–76 period. The blank rows separate different cross-barrier regions defined in the text: low-windward side, windward slope, immediate leeward slope, and low-lee side, respectively (from top to bottom).
The detailed topography, shown in Fig. 1, illustrates the complexity of the terrain, with low areas in the valleys and canyons where rivers flow. The Andes are roughly an ideal two-dimensional obstacle because of their perpendicular orientation to the westerlies, their narrow width (~150 km), and their high mean altitude (~4000 m), which, together with their proximity to the South Pacific Ocean (~200 km from the coast to the crest), make them an excellent barrier for moisture advection and precipitation coming from the west. The along-crest section in Fig. 2a highlights their high altitude, with three peaks above 6000 m altitude, the few passages below 3000 m north of 35°S, and the marked drop in height south of ~35°S. Thus, it is well known that the SCA profoundly alters the precipitation from midlatitude cyclones, although in a broad sense, through the climatic differences in adjacent low lands.
(a) Along-barrier height of the Andes at approximately the longitudes with higher altitudes and (b) mean cross-barrier height of the Andes between 32° and 37°S [topography data from U.S. Geological Service (USGS) with a 1-km horizontal resolution]. The large-scale cross-barrier areas are defined in the text using the few mountain stations available.
Citation: Journal of Hydrometeorology 12, 4; 10.1175/2010JHM1284.1
(a) Along-barrier height of the Andes at approximately the longitudes with higher altitudes and (b) mean cross-barrier height of the Andes between 32° and 37°S [topography data from U.S. Geological Service (USGS) with a 1-km horizontal resolution]. The large-scale cross-barrier areas are defined in the text using the few mountain stations available.
Citation: Journal of Hydrometeorology 12, 4; 10.1175/2010JHM1284.1
(a) Along-barrier height of the Andes at approximately the longitudes with higher altitudes and (b) mean cross-barrier height of the Andes between 32° and 37°S [topography data from U.S. Geological Service (USGS) with a 1-km horizontal resolution]. The large-scale cross-barrier areas are defined in the text using the few mountain stations available.
Citation: Journal of Hydrometeorology 12, 4; 10.1175/2010JHM1284.1
To estimate the large-scale variation in winter mean precipitation across the Andes, four main cross-barrier zones were established according the mean altitudes of Fig. 2b and the data availability: low-windward side, windward slope, immediate leeward slope, and low-leeward side. The low-windward side is the zone between the Pacific coast and the foot of the Andes with elevations less than 1000 m, the windward slope is the upslope zone of the barrier with elevations between 1500 and 3000 m, the immediate leeward slope is near the crest to the east of highest peaks (2000–3000 m), and the low-leeward side is the adjacent low-laying zone in Argentina with elevations less than 1500 m.
b. Spatial interpolation of precipitation and selection of heavy precipitation events for composites
A total of 78 stations with less than 5% missing values were considered to generate the continuous precipitation field in Fig. 3. Given the scarcity of stations, especially in the mountains, over a large area (~3.5 × 105 km2), we perform a simple spatial interpolation of precipitation. The successful application of more advanced approaches that incorporate the elevation in the terrain requires a high-density rain gauge network in order to derive a well-constrained precipitation–elevation relationship at the scale of the area of interest (Philips et al. 1992). Consequently, we used the kriging univariate interpolation method to estimate precipitation at non sampled sites (Cressie 1993). Kriging considers the information provided by surrounding sites through a pattern of spatial dependence between the observations (or roughness as characterized by variogram), which is typically produced by the terrain on precipitation. In fact, kriging has shown much better estimation of precipitation fields for low-density networks over a rough terrain than conventional Thiessen polygon and inverse square distance methods that ignore the pattern of spatial dependence (Goovaerts 2000).
(a) Spatial distribution of winter (Apr–Sep) mean precipitation (mm) for 1970–76. The values plotted are the winter mean precipitation at each station location, while the continuous lines represent the isohyet field obtained from observations by the kriging interpolation method. (b) Wintertime daily precipitation frequency (%) at windward-side (diamond), high-mountain (empty circles), and leeside (filled circles) stations.
Citation: Journal of Hydrometeorology 12, 4; 10.1175/2010JHM1284.1
(a) Spatial distribution of winter (Apr–Sep) mean precipitation (mm) for 1970–76. The values plotted are the winter mean precipitation at each station location, while the continuous lines represent the isohyet field obtained from observations by the kriging interpolation method. (b) Wintertime daily precipitation frequency (%) at windward-side (diamond), high-mountain (empty circles), and leeside (filled circles) stations.
Citation: Journal of Hydrometeorology 12, 4; 10.1175/2010JHM1284.1
(a) Spatial distribution of winter (Apr–Sep) mean precipitation (mm) for 1970–76. The values plotted are the winter mean precipitation at each station location, while the continuous lines represent the isohyet field obtained from observations by the kriging interpolation method. (b) Wintertime daily precipitation frequency (%) at windward-side (diamond), high-mountain (empty circles), and leeside (filled circles) stations.
Citation: Journal of Hydrometeorology 12, 4; 10.1175/2010JHM1284.1
As we will show in section 3b, heavy precipitation days (i.e., into the fourth quartile) accounts for most of the winter total precipitation at mountain stations (70%–80%; see Table 5). A heavy precipitation day is usually accompanied by consecutive light and/or other heavy precipitation days, resulting in a heavy precipitation event. These events have strong effects since they are one of the sources of water; and can also cause floods and landslides that block the mountain roads, producing great damage and economic losses. For this reason, we considered it appropriate to first focus on the climatology of heavy precipitation events.
Forty-six heavy events were selected from the daily precipitation data during the 7-yr period 1970–76. Daily precipitation amounts at the Lagunitas (LAG) and Puente del Inca (PIN) sites were averaged to generate the single LAG–PIN series. Although the precipitation amount in LAG is usually greater than in PIN, it typically occurs simultaneously at both sites, especially during heavy snow events. These two mountain stations were chosen because they are located at nearly the same latitude of the upstream rawinsonde launch (Fig. 1) and along the main highway between Chile and Argentina, so the synoptic climatology of heavy events would better contribute to forecasting purposes in this particular sector.
To create the 46-event dataset, we first retained a set of 94 heavy precipitation days of the LAG–PIN series. Then, the onset and end of each heavy precipitation event was determined by the previous and subsequent days with light precipitation around the day (or consecutive days) with heavy precipitation. A typical heavy precipitation event had a time scale of 2–4 precipitation days (1–2 light plus 1–2 heavy precipitation days; see Table 2). However, for periods between 7 and 12 consecutive rainy days (but not necessarily heavy consecutive days), we subdivided by successive heavy events, associating each subgroup of rainy days around 1 or more consecutive heavy rainy days with the passage of a frontal system and/or upper-level trough in the synoptic charts. Of course, this criterion for establishing the length of heavy events has limitations, but the more adequate satellite or hourly precipitation data were not available.
Frequency of heavy precipitation events observed at LAG–PIN series for different length of events (days) during the 7 winters from 1970 to 1976. For starting and ending day of each event see the first column of Table 6.
To verify that the daily precipitation series from LAG–PIN is representative of the precipitation in this Andean region, it was compared with the number of heavy events generated by the daily series from Valle Hermoso (~35°S). The Valle Hermoso series determined only four events more than the ones obtained from the LAG–PIN series—75% of which coincided in both series. This comparison also reflects that most of the heavy events cover the entire study region because of the synoptic scale of the phenomenon.
c. Composite of mean and anomaly synoptic fields and vertical profiles
The synoptic-scale pattern associated with heavy orographic precipitation events were identified by composites of the mean and standardized anomaly fields using the four-times daily 40-yr European Centre for Medium-Range Weather Forecasts (ECMWF) Re-Analysis (ERA-40) data (Uppala et al. 2005), which is available in a 2.5° × 2.5° grid at 0000, 0600, 1200, and 1800 UTC. We identified synoptic-scale patterns by separating the heavy precipitation events into 2 subcategories: intense and extreme events. Intense and extreme events are defined as daily precipitation is within percentiles 75%–95% and 95%–100%, respectively. Useful information to infer synoptic forcing on orographic precipitation could arise from differences in the meteorological variables between these two cases.
The standardized anomalies method has been widely used to characterize synoptic-scale precipitation events, which then serves to help forecasters in their pattern recognition (e.g., Grumm and Hart 2001; Stuart and Grumm 2006; Junker et al. 2008), as well as to categorize objectively extreme precipitation events (Hart and Grumm 2001; Graham and Grumm 2010). As applied in previous studies, the departure of the meteorological variable at each grid point from its 21-day running climatological mean is divided by its 21-day running standard deviation according to the following equation: 
The composite vertical profiles were calculated using the rawinsonde data at Quinteros (QUI) on the Chilean coast (Fig. 1). The data and their metadata were obtained from the Integrated Global Radiosonde Archive (IGRA) and the rawinsonde dataset from the National Climatic Data Center (NCDC). A complete description of the dataset and details of the quality control are given by Durre et al. (2006). To calculate the composite vertical profiles, we used the sounding launched at 0000 UTC of the maximum precipitation day of each event, since that time is the middle of the daily precipitation observation period and, thus increases the probability of being close to subperiods of maximum or total daily precipitation. Moreover, selecting only the day of maximum precipitation ensures more independence among events (mainly in consecutive rainfall events). In general, the vertical resolution of data is good—most of them having more than 20 vertical levels from surface to 300 hPa. The soundings of each event have the standard levels and some intermediate levels, which differed among observations. The most frequent intermediate levels were defined to homogenize all the profiles, totaling 19 vertical levels between 1000 and 300 hPa. During days with no observations made exactly at these established intermediate levels, linear interpolations were applied between continuous levels.
3. Climatology of winter orographic precipitation over the subtropical central Andes
This section provides additional details of winter precipitation over the SCA related to previous studies covering Argentina, Chile, and the South American continent (Hoffman 1975; Prohaska 1976; Miller 1976) as well as expands previous analysis limited to 33°S latitude (Ereño and Hoffman 1978) and to the windward slope between 32° and 35°S (FG07). Specifically, this paper expands the analysis to what occurs in the immediate leeward slopes, even within the high-mountain region, as well as in the low-leeward side.
a. Spatial distribution of mean winter precipitation over the Andes and surrounding lowlands
Two main features stand out from the mean winter precipitation field in Fig. 3a: the marked northward decay in the low-windward rainfall and the strong east–west gradient in the Andes to the south of 32°S. In the low-windward side, winter precipitation ranges from ~800 mm at 36°S to less than 100 mm at 30°S. This feature has been widely documented by Miller (1976), Montecinos et al. (2000), and FG07, among others, who indicate that central Chile is a transition zone between the permanent arid climate in the north of Chile (one of the driest in the world) and the humid midlatitude region with extratropical precipitation. On the other hand, isohyets change from a north–south gradient in the low-windward side to a strong east–west gradient in the Andes to the south of 32°S. Although few mountain stations were available, they generated small local precipitation maxima on the windward slopes in the mean precipitation field around 33°–34° and 36°S (i.e., below the crest of the Andes, which is represented approximately by the Argentina–Chile border in Fig. 3a). After reaching these maxima on windward slopes, the precipitation amounts drop markedly toward the east, with values in low-lee side representing at most 10% of the maxima observed on windward slopes.
The daily winter precipitation frequency was lower on the low-lee than on the low-windward side and high mountains (Fig. 3b). In general, the northward decrease in frequency observed at all stations, independent of their location (Fig. 3b), could partly explain the northward decrease in the amount of precipitation across the study area (Fig. 3a). This feature is in agreement with the winter precipitation on the windward side shown by FG07.
The cross-barrier precipitation plots in Fig. 4 highlight the profound alteration that precipitation systems in this sector of the Andes undergo. In the central-latitude bands (Figs. 4b,c and Table 3), the observations record an increase in precipitation with height, reaching a maximum on the windward slope (~700 mm), which decreases for stations in the immediate leeward slope, ~(300–400) mm. The orographic enhancement was close to 2.5 and 1.5 around 33° and 34°S (Table 3), which is similar to that reported by FG07 on the windward slope. Similarly, the reduction rates at these latitudes were around 2.5 and 1.6 on the immediate leeward slope (Table 3). The observations also recorded notable reductions of ~9 and ~6 times that of mean precipitation on the low-leeward side with respect to the mean low-windward-side precipitation in 33° and 34°S bands, respectively (Table 3). These enhancements and reductions in precipitation in the cross-mountain direction were also appreciable in the southernmost latitudinal band (Fig. 4d), although to a slightly lesser degree (Table 3). In the most-sampled band around 33°S (Fig. 4b), the abrupt changes of mean precipitation amount at high-mountain stations was recorded with greater detail, and is also denoted by the high standard deviation on the immediate leeward slope in Table 3. It is also noted in Table 3 and Fig. 4 that the precipitation amount on the immediate leeward slope is similar to that on the low-windward side, and the rain shadow effect on the low-lee side is accentuated northward, which is consistent with the northward increase of the mountain range altitude (Fig. 2a).
Cross-barrier plots of mean winter precipitation for different latitudinal bands: (a) 31.15°–32.45°S, (b) 32.45°–33.50°S, (c) 33.50°–34.50°S, and (d) 34.60°–36.5°S. Topographic profiles correspond to the middle of each band, and were plotted using USGS topographic data with observations enclosed in the horizontal square across the Andes (Fig. 1) adjusted (see text).
Citation: Journal of Hydrometeorology 12, 4; 10.1175/2010JHM1284.1
Cross-barrier plots of mean winter precipitation for different latitudinal bands: (a) 31.15°–32.45°S, (b) 32.45°–33.50°S, (c) 33.50°–34.50°S, and (d) 34.60°–36.5°S. Topographic profiles correspond to the middle of each band, and were plotted using USGS topographic data with observations enclosed in the horizontal square across the Andes (Fig. 1) adjusted (see text).
Citation: Journal of Hydrometeorology 12, 4; 10.1175/2010JHM1284.1
Cross-barrier plots of mean winter precipitation for different latitudinal bands: (a) 31.15°–32.45°S, (b) 32.45°–33.50°S, (c) 33.50°–34.50°S, and (d) 34.60°–36.5°S. Topographic profiles correspond to the middle of each band, and were plotted using USGS topographic data with observations enclosed in the horizontal square across the Andes (Fig. 1) adjusted (see text).
Citation: Journal of Hydrometeorology 12, 4; 10.1175/2010JHM1284.1
Average precipitation for different cross-barrier zones and latitudes (bold; see text for definition). Increase and reduction rates for each latitudinal band of Fig. 4. The numbers in square brackets and parentheses represent the standard deviation (σ) and available stations (used to calculate the average and σ of each zone), respectively (see also Figs. 1 and 2b).
b. Different precipitation regimes across the subtropical central Andes
In the broad sense, the existence of different precipitation regimes in the low-windward and low-leeward side has been mentioned in large-scale climatic studies (e.g., Prohaska 1976; Miller 1976; Compagnucci and Vargas 1998; Montecinos et al. 2000; Masiokas et al. 2006). Nonetheless, we take advantage of the few additional stations available in the mountain and surroundings to examine the different precipitation regimes across the SCA.
In the period 1970–76, winter precipitation was above 80% of the annual total at all stations in the low-windward side, windward slope, and immediate leeward slope, reaching 100% of annual precipitation in some years and stations (not shown). In contrast, the winter total did not exceed 40% of the annual total at stations in the low-leeward side north of 35°S, indicating a summer precipitation regime only for this cross-barrier zone. In the low-lee side south of 35°S, the winter precipitation represented about 40%–60% of the annual total at Malargüe Aero (MAL) and Buta Ranquil (BRQ) stations (see Table 1 and Fig. 1a). The 1961–90 mean for MAL and BRQ shows two peaks: one in summer and the other in winter. These stations are located in the low-leeside transition zone, between the low lee of the high subtropical Andes with summer convective rain regime and the low lee of the Patagonian Andes with winter spillover precipitation regime.
Corrected lineal correlation matrix (
Days with precipitation in the fourth quartile account for nearly 75% of the winter total over the low-windward side, windward, and immediate leeward slopes (Table 5). This means that most of the winter precipitation in these mountain zones, which is almost equal to the annual total, occurs on the heaviest precipitation days, with a frequency of only 25%. Moreover, the extreme precipitation days (into the 95%–100% percentile) account for between 30% and 35% of the winter total at most sites (Table 5), with an approximate frequency of two days per year. These features in the daily precipitation distribution were independent of the altitude, suggesting that they are related to latitude rather than topography.
Basic statistics of winter precipitation series for different percentiles and the total (tot) of each series. Frequency (days during the period 1970–76), winter mean (mm winter−1), daily mean (mm day−1), and percentage of accumulated rainfall for each percentile with respect to tot are shown. Stations are QTN, LAG, VPA, PIN, VHE, and MAL (see Table 1).
4. Synoptic climatology of the heavy precipitation events over the subtropical central Andes
Based on the fact that most precipitation accumulated in the fourth quartile (section 3b), here we identify the synoptic-scale patterns of the heavy precipitation events subdivided into intense and extreme events to infer synoptic forcing on orographic precipitation.
a. Synoptic-scale patterns during the intense and extreme orographic precipitation events
The composite fields revealed that surface low pressure systems on the Pacific coast of South America accompanied the heavy orographic precipitation events, which were stronger for the extreme rather than just the intense events (Figs. 5a–c). The subtropical high and the post-cold-front ridge were also stronger at the surface off the Chilean coast for such extreme events. Consequently, northwesterly flow at the surface upstream of the Andes was also greater during the extreme rather than intense events (Figs. 5a,b). The fact that deeper surface midlatitude cyclones on the Pacific coast of South America accompanied these extreme than the intense events was confirmed by the statistical test in Fig. 5d. The Student t test (Fisher) revealed that the mean (standard deviation) of sea level pressure (SLP) for the group of extreme and intense events was (was not) significantly different. Thus, the difference in the mean SLP (Fig. 5c) would be explained by a difference in the intensity of surface cyclones between each group of events, rather than a difference in the spatial dispersion of cyclones (related to the spatial overlap of cyclones). The features observed in the baroclinic waves at the surface remained from low to high levels (not shown), with greater moisture flux and winds for extreme than intense events.
Composite of SLP (hPa; contours every 3 hPa) and winds (m s−1; half barb = 5 m s−1, full barb = 10 m s−1, and magnitude shaded) for (a) extreme and (b) intense events. (c) Difference in SLP between extreme and intense events (contours) with shaded areas corresponding to significant values at 90%, 95%, 97.5%, and 99% levels (from light to dark gray) for different SLP mean between extreme and intense events, according to the Student t test. (d) Shaded areas correspond to the significant values at same levels as (c) but for different std devs of SLP between extreme and intense events, according to the Fisher statistical test.
Citation: Journal of Hydrometeorology 12, 4; 10.1175/2010JHM1284.1
Composite of SLP (hPa; contours every 3 hPa) and winds (m s−1; half barb = 5 m s−1, full barb = 10 m s−1, and magnitude shaded) for (a) extreme and (b) intense events. (c) Difference in SLP between extreme and intense events (contours) with shaded areas corresponding to significant values at 90%, 95%, 97.5%, and 99% levels (from light to dark gray) for different SLP mean between extreme and intense events, according to the Student t test. (d) Shaded areas correspond to the significant values at same levels as (c) but for different std devs of SLP between extreme and intense events, according to the Fisher statistical test.
Citation: Journal of Hydrometeorology 12, 4; 10.1175/2010JHM1284.1
Composite of SLP (hPa; contours every 3 hPa) and winds (m s−1; half barb = 5 m s−1, full barb = 10 m s−1, and magnitude shaded) for (a) extreme and (b) intense events. (c) Difference in SLP between extreme and intense events (contours) with shaded areas corresponding to significant values at 90%, 95%, 97.5%, and 99% levels (from light to dark gray) for different SLP mean between extreme and intense events, according to the Student t test. (d) Shaded areas correspond to the significant values at same levels as (c) but for different std devs of SLP between extreme and intense events, according to the Fisher statistical test.
Citation: Journal of Hydrometeorology 12, 4; 10.1175/2010JHM1284.1
The standardized anomaly fields at low levels emphasize the difference in the magnitude of the low pressure system and moisture flux between extreme and intense events (Fig. 6). Even though both cases showed negative geopotential height and positive moisture flux anomalies off the Pacific coast of South America on the previous day (day −1; Fig. 6, left) to the highest 24-h precipitation rate for each event (day 0), the extreme composite exhibited a stronger strengthening of a landfalling extratropical cyclone than the intense composite on day 0 (right panels in Fig. 6). The higher anomalies of both variables were around (0.3–0.6)σ on day 0 for intense events, whereas extreme events exceed 1.2σ. In general, the anomalous northwesterly moisture flux was slightly stronger at 700 hPa than at 850 hPa on day 0 (cf. Figs. 6b–d,f–h), which is consistent with the strongest statistical relationship between the moisture flux at 2500 m and the winter rain in central Chile found by FG07. In addition, the northwesterly flow upstream of the Andes and the position of lower geopotential height anomalies indicate that highest precipitation rates typically occurred within the warm sector of the cyclone that is in the pre-cold-front environment.
Composite of normalized 850-hPa geopotential height anomaly (contours) and moisture flux (shaded; vectors indicate direction) at (left) day −1 and (right) day 0 for (a),(b) extreme and (c),(d) intense events. (e)–(h) As in (a)–(d), but at 700-hPa level. Contours and shaded intervals of normalized anomalies are 0.3σ.
Citation: Journal of Hydrometeorology 12, 4; 10.1175/2010JHM1284.1
Composite of normalized 850-hPa geopotential height anomaly (contours) and moisture flux (shaded; vectors indicate direction) at (left) day −1 and (right) day 0 for (a),(b) extreme and (c),(d) intense events. (e)–(h) As in (a)–(d), but at 700-hPa level. Contours and shaded intervals of normalized anomalies are 0.3σ.
Citation: Journal of Hydrometeorology 12, 4; 10.1175/2010JHM1284.1
Composite of normalized 850-hPa geopotential height anomaly (contours) and moisture flux (shaded; vectors indicate direction) at (left) day −1 and (right) day 0 for (a),(b) extreme and (c),(d) intense events. (e)–(h) As in (a)–(d), but at 700-hPa level. Contours and shaded intervals of normalized anomalies are 0.3σ.
Citation: Journal of Hydrometeorology 12, 4; 10.1175/2010JHM1284.1
At middle and upper levels (Fig. 7), the negative geopotential height and westerly wind anomalies also moved eastward from day −1 to day 0 off the Pacific coast, with much stronger magnitudes for extreme than intense cases. Minimum values of the height anomalies tilted to the west with increasing altitude, which indicates the baroclinic character of the low pressure systems (Figs. 6 and 7). Moreover, the negative and positive geopotential height anomalies determined a dipole ridge–trough pattern over the southeastern Pacific Ocean—more evident for extreme than intense events (Fig. 7). A similar dipole synoptic pattern enhancing the flow toward the Andes was also observed by VN09 for a heavy winter precipitation case study. Furthermore, this dipole pattern presents some resemblance to those observed for heavy precipitation composites over the Sierra Nevada (Junker et al. 2008; Pandey et al. 1999) and in the Pacific Northwest (Lackmann and Gyakum 1999), but it differs from the one over the South Pacific with regard to the position of a positive height anomaly located southeast (downstream) of the negative counterpart for the precipitation cases of North America.
As in Fig. 6, but for mean normalized anomalies of 500- and 250-hPa geopotential height (contours) and winds (shading indicates magnitude and arrows indicate direction). Contours and shading intervals are 0.3σ.
Citation: Journal of Hydrometeorology 12, 4; 10.1175/2010JHM1284.1
As in Fig. 6, but for mean normalized anomalies of 500- and 250-hPa geopotential height (contours) and winds (shading indicates magnitude and arrows indicate direction). Contours and shading intervals are 0.3σ.
Citation: Journal of Hydrometeorology 12, 4; 10.1175/2010JHM1284.1
As in Fig. 6, but for mean normalized anomalies of 500- and 250-hPa geopotential height (contours) and winds (shading indicates magnitude and arrows indicate direction). Contours and shading intervals are 0.3σ.
Citation: Journal of Hydrometeorology 12, 4; 10.1175/2010JHM1284.1
The specific humidity anomalies were positive on the Chilean coast, increasing in magnitude from surface to 700 hPa and being stronger for extreme than intense events (Fig. 8). On the other hand, the temperature anomaly fields showed positive values in subtropical latitudes far away from the Chilean coast and negative values in midlatitude close to the Chilean coast—these also being higher for extreme than intense episodes (Fig. 8). The positive temperature anomalies indicate the deepening of the subtropical anticyclone, while the negative anomalies indicate an extratropical cyclone approaching the Chilean coast. Surprisingly, there was not a positive–negative dipole pattern in the temperature anomalies around the cyclone as observed in previous composite studies observed along the Pacific North American coast (Lackmann and Gyakum 1999; Pandey et al. 1999; Junker et al. 2008; Neiman et al. 2008). These previous studies used precipitation gauges located on higher latitudes (north of 37°N) than ones used in our study (32.5°–33°S) to perform the composite fields of precipitation events. Accordingly, the cyclone environment along the western coast of North America may be less affected by the subtropical anticyclone circulation than the cyclone environment of precipitation cases analyzed here, where the typical descending and relatively warmer air from the South Pacific anticyclone circulation (closer to the continent in winter; Satyamurty et al. 1999) would be replaced by ascending cooler air while cold fronts and coupled-troughs aloft approached the coast of South America.
Composite normalized temperature anomaly field (contours) and specific humidity (shading) at (a),(b) surface, (c),(d) 850-, and (e),(f) 700-hPa levels for (a),(c),(e) extreme and (b),(d),(f) intense events. Contours and shading of normalized anomalies are 0.3σ.
Citation: Journal of Hydrometeorology 12, 4; 10.1175/2010JHM1284.1
Composite normalized temperature anomaly field (contours) and specific humidity (shading) at (a),(b) surface, (c),(d) 850-, and (e),(f) 700-hPa levels for (a),(c),(e) extreme and (b),(d),(f) intense events. Contours and shading of normalized anomalies are 0.3σ.
Citation: Journal of Hydrometeorology 12, 4; 10.1175/2010JHM1284.1
Composite normalized temperature anomaly field (contours) and specific humidity (shading) at (a),(b) surface, (c),(d) 850-, and (e),(f) 700-hPa levels for (a),(c),(e) extreme and (b),(d),(f) intense events. Contours and shading of normalized anomalies are 0.3σ.
Citation: Journal of Hydrometeorology 12, 4; 10.1175/2010JHM1284.1
b. Regional atmospheric conditions during intense and extreme precipitation events
Wind distribution at different vertical levels during the total set of 94 days with heavy precipitation is shown in Fig. 9. The wind rose built with rawinsonde data shows that the flow east of the Andes is mostly from the north at near-surface levels and that a westerly component appears and enhances with increasing height. Such flow distribution is coherent with pre-cold-front surface conditions, a coupled upper-level trough, and a predominant northerly component at lower levels, which suggests an orographic blocking. For near surface, and to a lesser extent at higher levels, the southerly component was infrequent, which indicates the lack of precipitation days in post-cold-front situations.
Wind roses for (a) 1000, (b) 850, (c) 700, and (d) 500 hPa using the rawinsonde data from QUI, located on the windward side of the Andes, during the days classified as extreme and intense.
Citation: Journal of Hydrometeorology 12, 4; 10.1175/2010JHM1284.1
Wind roses for (a) 1000, (b) 850, (c) 700, and (d) 500 hPa using the rawinsonde data from QUI, located on the windward side of the Andes, during the days classified as extreme and intense.
Citation: Journal of Hydrometeorology 12, 4; 10.1175/2010JHM1284.1
Wind roses for (a) 1000, (b) 850, (c) 700, and (d) 500 hPa using the rawinsonde data from QUI, located on the windward side of the Andes, during the days classified as extreme and intense.
Citation: Journal of Hydrometeorology 12, 4; 10.1175/2010JHM1284.1
Mean vertical profiles of extreme and intense cases are shown in Fig. 10. The cross-barrier winds were from the west with magnitudes above the climatological mean profile at all levels and in both cases, but greater in extreme episodes than in intense ones (Fig. 10a). In the 800–600-hPa layer, the mean westerly flow for extreme events departed 1.5σ from the climatological mean, which would be consistent with a greater upslope ascent of the cross-barrier flow. On the other hand, the along-barrier winds were poleward at all levels and also stronger for extreme than intense episodes (Fig. 10b). In both composite mean profiles, the poleward wind peaks at about 700 hPa, suggesting the development of barrier jets that are typical when flow is blocked by the topography (e.g., Parish 1982; Marwitz 1987; McCauley and Sturman 1999; Garvet et al. 2007).
Mean vertical profiles of the (a) cross-barrier and (b) along-barrier wind components, (c) square of dry Brunt–Väisälä frequency (10−4 s−2), and (d) temperature for extreme (black solid lines) and intense (dashed black lines) episodes. The profiles were calculated using the rawinsonde observations from QUI in the central Chilean coast. Winter (Apr–Sep) 1961–90 climatological mean (full lines) and the std dev (dashed gray lines) are shown.
Citation: Journal of Hydrometeorology 12, 4; 10.1175/2010JHM1284.1
Mean vertical profiles of the (a) cross-barrier and (b) along-barrier wind components, (c) square of dry Brunt–Väisälä frequency (10−4 s−2), and (d) temperature for extreme (black solid lines) and intense (dashed black lines) episodes. The profiles were calculated using the rawinsonde observations from QUI in the central Chilean coast. Winter (Apr–Sep) 1961–90 climatological mean (full lines) and the std dev (dashed gray lines) are shown.
Citation: Journal of Hydrometeorology 12, 4; 10.1175/2010JHM1284.1
Mean vertical profiles of the (a) cross-barrier and (b) along-barrier wind components, (c) square of dry Brunt–Väisälä frequency (10−4 s−2), and (d) temperature for extreme (black solid lines) and intense (dashed black lines) episodes. The profiles were calculated using the rawinsonde observations from QUI in the central Chilean coast. Winter (Apr–Sep) 1961–90 climatological mean (full lines) and the std dev (dashed gray lines) are shown.
Citation: Journal of Hydrometeorology 12, 4; 10.1175/2010JHM1284.1
The along-barrier jet development at low levels would be coherent with a stable mean profile (Fig. 10c), probably because of the strong influence of the southeastern Pacific subtropical anticyclone. This condition determines a blocked flow regime at low levels, with a typical Froude number (Fr) less than 0.5 from 750 hPa to surface (i.e., Fr = u/hN, where according to Fig. 10, the zonal wind u = 10 m s−1, the dry Brunt–Väisälä frequency N = 1 × 10−2 s−1, and a height h = 2000 m from 750 hPa to crest level). Meanwhile, the along-barrier jet at 700 hPa could respond to dynamic blocking on the upslope below the range crest, as well as to the diabatic process of melting around the 0°C isotherm (see Fig. 10d) proposed by Marwitz (1987). Along-barrier jets were also observed during the heavy precipitation case study over the SCA by VN09. Finally, the temperature mean profile showed slight cold anomalies from 950 to 400 hPa during both composite events (Fig. 10d), confirming the below-normal temperatures upstream of the Andes recently depicted by reanalysis data.
The descendent nature of the flow inhibiting precipitation on the lee side during heavy precipitation events in the Andes is inferred by the recurrent downslope wind extending fully to the low-leeward side (Table 6). Downslope wind (named “zonda” in South America; Norte et al. 2008; Seluchi et al. 2003) and associated föehn effects (i.e., a sharply increased dewpoint depression) were recorded north of 34°S by at least one (two) low-lee station(s) in 80% (67%) of 46 total heavy precipitation events. In addition, regional features can be noticed in Table 6 as the higher frequency of zonda wind on the northernmost leeside 30°–32°S latitude band, rather than the 32°–34°S band, in agreement with Norte (1988). Dewpoint depressions commonly exceeded 30°C, and strong northwest–west winds occurred in most cases (Table 6), denoting the abrupt drying of the air mass by airflow subsidence. Within the 34°–36°S latitude band, the MAL station recorded a high frequency of zonda winds, but in general there were lower dewpoint depressions than those recorded farther north (see Table 6). These differences in dewpoint depression may derive from a shorter downslope track of air parcels near MAL than from ones farther north because of both the higher altitude of MAL (~1450 m) compared to the remaining low-laying lee sites and the lower elevation of the Andes in the southernmost latitude band (Fig. 2a). Additionally, the proximity of MAL to the foot of the Andes and its southernmost location (closer to the storm track core) would explain the slightly different evolution of heavy storms at MAL from sites farther north. MAL usually records spillover precipitation interrupted by zonda wind periods during heavy storms in the Andes, whereas lee sites farther north are strongly influenced by downslope flow, which is denoted by the near absence of spillover precipitation (e.g., see Fig. 6 of VN09). As a result, the winter precipitation from midlatitude cyclones is largely inhibited on the northern low lee and less inhibited on the southern low lee of SCA, as reflected in the winter climatology in section 3.
Low-leeward side stations that recorded downslope winds (zonda wind) within 30°–32°S, 32°–34°S, and 34°–36°S lat bands during the extreme and intense precipitation episodes over the subtropical central Andes. Stations are BAL, JAC, SJA, MZA, SCA, SRA, and MAL (see Fig. 1 and Table 1). Local time is (UTC −3 h), so temperature > 15°C during night and morning in winter would be an indicator of föehn effect.
c. Association between intense and extreme events and atmospheric rivers
Previous studies have shown the significant hydrometeorological role of atmospheric rivers along the western coast of North America in providing water and melting snowpack (Neiman et al. 2008; Guan et al. 2010) and in producing flooding and landslides in low-lying areas adjacent to the mountain ranges (Colle and Mass 2000; Ralph et al. 2005a, 2006; among others). Based upon the similarities between the western Pacific coasts of North and South America, we investigate the possible association of heavy orographic precipitation events with landfalling atmospheric rivers.
The criterion used to detect atmospheric rivers follows the method proposed by Ralph et al. (2004), which uses the integrated water vapor (IWV) from Special Sensor Microwave Imager (SSM/I) satellite data as a detection proxy. SSM/I and some other satellite data were not available for 1970–76, so we used the ERA-40 reanalyses to identify long (>2000 km) and narrow (<1000 km) IWV plumes as proxies of atmospheric rivers. The NCEP–NCAR reanalyses (Kalnay et al. 1996), with the same 2.5° × 2.5° resolution as that for ERA-40, represented somewhat coarser IWV plumes than SSM/I data over the Pacific coast of North America; however, they accurately depict the position and orientation of the IWV plumes (Neiman et al. 2008). Based on these previous reliable results of the reanalyses data, we used ERA-40 data to examine the large-scale conditions of atmospheric rivers. To favor a greater spatial case-to-case overlap in the compositing of the narrow and fast-moving IWV plumes, we average the IWV values in the 12-h period of day with maximum 24-h precipitation in each event, when the IWV plume landfall was closer to the latitude of the LAG and PIN sites (~33°S). Additionally, to confirm that IWV plumes are related to vapor transport toward the Andes, we evaluated the composite of vertically integrated horizontal water vapor transport (IVT),2 calculated with ERA-40 data as in Neiman et al. (2008, p. 31).
The composite of extreme events showed a narrow elongated IWV plume extending toward the coast of South America, whose maximum ranges from 20 to 23 mm (Fig. 11a). For intense events, the IWV plume was somewhat wider (Fig. 11b). Large-scale conditions exhibited a midlatitude belt of IVT from the Pacific Ocean to the Andes that was stronger in extreme than in intense cases (cf. Figs. 11c,d). On the Chilean coast, IVT reached 250–350 kg m−1 s−1 in the extreme events (Fig. 11c), whereas IVT reached 200–250 kg m−1 s−1 in intense events (Fig. 11d). The standardized IVT anomalies highlight the differences in intensity, orientation, and extension of the moisture transport belt among case composites, being stronger, longer, and more cross-mountain oriented in extreme than intense episodes (Figs. 11e,f). Figure 11 confirmed that the IWV plumes in extreme cases were in fact associated with strong westerly-northwesterly IVTs, which impinge on the Andes. On the other hand, the IWV mean composites of intense cases did not show a filamentary structure; even so, most of those cases were associated with a IWV plume and anomalous vapor transport to the Andes.
Composite mean of (a),(b) precipitable water or IWV (mm; shading), (c),(d) IVT (kg m−1 s−1; magnitude shaded), and (e),(f) normalized anomalies of IVT (magnitude shaded every 0.3σ) for (a),(c),(e) extreme and (b),(d),(f) intense episodes.
Citation: Journal of Hydrometeorology 12, 4; 10.1175/2010JHM1284.1
Composite mean of (a),(b) precipitable water or IWV (mm; shading), (c),(d) IVT (kg m−1 s−1; magnitude shaded), and (e),(f) normalized anomalies of IVT (magnitude shaded every 0.3σ) for (a),(c),(e) extreme and (b),(d),(f) intense episodes.
Citation: Journal of Hydrometeorology 12, 4; 10.1175/2010JHM1284.1
Composite mean of (a),(b) precipitable water or IWV (mm; shading), (c),(d) IVT (kg m−1 s−1; magnitude shaded), and (e),(f) normalized anomalies of IVT (magnitude shaded every 0.3σ) for (a),(c),(e) extreme and (b),(d),(f) intense episodes.
Citation: Journal of Hydrometeorology 12, 4; 10.1175/2010JHM1284.1
The low-level jet structure of atmospheric rivers is suggested upstream of the Andes during the 46 highest precipitation events by the rawinsonde data (Fig. 12a). When the low-level u component of the wind was at a maximum within the 850–700-hPa layer, the west wind velocity exceeded 2σ from the 1961–90 winter mean in most heavy precipitation cases, whereas the low-level u wind maximum in the surface–850-hPa layer showed west velocity exceeding 2σ in all cases. A similar domain of strong anomalies in poleward winds was recorded when the υ component was at a maximum below 700 hPa (Fig. 12a), suggesting the recurrence of a low-level jet within the pre-cold-front sector of cyclones during heavy orographic precipitation. In addition, a high value of IWV was represented by ERA-40 reanalysis over the Chilean coast (Fig. 12b), usually surpassing the threshold proposed by Ralph et al. (2006) to determine the arrival of an atmospheric river on the California coast, along with a low-level jet and IWV plume structure. An individual inspection of reanalysis charts for each precipitation event revealed that only one extreme and eight intense of 46 cases did not exhibit a long and narrow IWV plume. These nine events that were not linked to IWV plumes showed cut-off lows at upper levels during post-cold and warm front synoptic situations. Whereas the majority of events (37) related to IWV plumes occurred in the pre-cold-front environment just before of the cold front passage; supporting the fact that heavy precipitation on the Andes is linked to landfalling atmospheric rivers.
(a) Scatterplot of u and υ wind pair when u component (black filled symbols) and υ component (unfilled symbols) are at a maximum at some level below 700 hPa for the tot precipitation episodes. The circles and triangles represent cases when the u or υ component is at a maximum in the surface–850-hPa layer or in the 850–700-hPa layer, respectively. The wind was measured by the rawinsonde at QUI on the Chilean coast, and the vertical and horizontal dotted lines correspond to two winter 1961–90 std devs of the west and north components, respectively. (b) Max IWV on the continental coast (i.e., the max at grid points 32.5°S–72.5°W or 35°S–72.5°W of the ERA-40 reanalysis) for the 46 tot most intense wintertime (Apr–Sep) events during 1970–76. The horizontal dotted line corresponds to the threshold proposed by Ralph et al. (2004) to determine the landfalling of an atmospheric river on the U.S. west coast, along with the spatial structure of IWV plumes and the occurrence of strong low-level winds.
Citation: Journal of Hydrometeorology 12, 4; 10.1175/2010JHM1284.1
(a) Scatterplot of u and υ wind pair when u component (black filled symbols) and υ component (unfilled symbols) are at a maximum at some level below 700 hPa for the tot precipitation episodes. The circles and triangles represent cases when the u or υ component is at a maximum in the surface–850-hPa layer or in the 850–700-hPa layer, respectively. The wind was measured by the rawinsonde at QUI on the Chilean coast, and the vertical and horizontal dotted lines correspond to two winter 1961–90 std devs of the west and north components, respectively. (b) Max IWV on the continental coast (i.e., the max at grid points 32.5°S–72.5°W or 35°S–72.5°W of the ERA-40 reanalysis) for the 46 tot most intense wintertime (Apr–Sep) events during 1970–76. The horizontal dotted line corresponds to the threshold proposed by Ralph et al. (2004) to determine the landfalling of an atmospheric river on the U.S. west coast, along with the spatial structure of IWV plumes and the occurrence of strong low-level winds.
Citation: Journal of Hydrometeorology 12, 4; 10.1175/2010JHM1284.1
(a) Scatterplot of u and υ wind pair when u component (black filled symbols) and υ component (unfilled symbols) are at a maximum at some level below 700 hPa for the tot precipitation episodes. The circles and triangles represent cases when the u or υ component is at a maximum in the surface–850-hPa layer or in the 850–700-hPa layer, respectively. The wind was measured by the rawinsonde at QUI on the Chilean coast, and the vertical and horizontal dotted lines correspond to two winter 1961–90 std devs of the west and north components, respectively. (b) Max IWV on the continental coast (i.e., the max at grid points 32.5°S–72.5°W or 35°S–72.5°W of the ERA-40 reanalysis) for the 46 tot most intense wintertime (Apr–Sep) events during 1970–76. The horizontal dotted line corresponds to the threshold proposed by Ralph et al. (2004) to determine the landfalling of an atmospheric river on the U.S. west coast, along with the spatial structure of IWV plumes and the occurrence of strong low-level winds.
Citation: Journal of Hydrometeorology 12, 4; 10.1175/2010JHM1284.1
As an example, four precipitation events associated with atmospheric rivers are shown in Fig. 13. The four events exhibited long IWV plumes aligned with the cold front and stretching from the ocean to the western coast of South America (Fig. 13, left). The concentrated low-level moisture fluxes were represented by the ERA-40 data, which were stronger and oriented in a more cross-barrier direction in the two extreme than the two intense cases (Fig. 13, middle). The time-height plots at the 32.5°S–72.5°W grid point show the pre-frontal northwesterly flow during the four episodes (Fig. 13, right). Moreover, the first and/or only IWV maximum is reached hours prior to the cold front passage, which is identified by an enhancement of the westerly wind; meaning at 0600 UTC 8 May prior to the front passage at 0000 UTC 9 May 1972 (Fig. 13c), between 1200 and 1800 UTC 24 June before 1200 UTC 25 June 1974 (Fig. 13f), at 1800 UTC 5 July before 0060 UTC 6 July 1972 (Fig. 13i), and at 1800 UTC 20 May before 0000 UTC 21 May 1974 (Fig. 13l). Extreme cases 3 and 11 had the greatest impact among the 46 cases studied, causing emergency situations in central Chile with casualties, floods, river overflow, and snow avalanches. More recent examples of atmospheric rivers impinging on the SCA were observed by SSM/I satellite imagery during two heavy orographic precipitation case studies (Figs. 4.4 and 5.4 in Viale 2010).
(a),(d),(g),(j) SLP (contours every 4 hPa) and IWV (shading every 5 mm) fields. (b),(e),(h),(k) Cross sections of IWV plumes at 80°W showing the cross-barrier component and the tot magnitude of moisture flux (contours and shading every 30 m s1 g kg1, respectively), and (c),(f),(i),(l) time-altitude plots showing winds (barbs; half barb = 5 m s−1 and full barb = 10 m s−1), u wind (full lines every 5 m s1), and IWV (dashed-point line in mm) at 32.5°S–72.5°W grid point of ERA-40 for four heavy precipitation events. The extreme storms of (a)–(c) 5–8 May 1972 and (d)–(f) 24–30 Jun 1974. These storms caused flooding and great damages and casualties in central Chile. The intense storms of (g)–(i) 5–6 Jul 1972 and (j)–(l) 18–22 May 1974.
Citation: Journal of Hydrometeorology 12, 4; 10.1175/2010JHM1284.1
(a),(d),(g),(j) SLP (contours every 4 hPa) and IWV (shading every 5 mm) fields. (b),(e),(h),(k) Cross sections of IWV plumes at 80°W showing the cross-barrier component and the tot magnitude of moisture flux (contours and shading every 30 m s1 g kg1, respectively), and (c),(f),(i),(l) time-altitude plots showing winds (barbs; half barb = 5 m s−1 and full barb = 10 m s−1), u wind (full lines every 5 m s1), and IWV (dashed-point line in mm) at 32.5°S–72.5°W grid point of ERA-40 for four heavy precipitation events. The extreme storms of (a)–(c) 5–8 May 1972 and (d)–(f) 24–30 Jun 1974. These storms caused flooding and great damages and casualties in central Chile. The intense storms of (g)–(i) 5–6 Jul 1972 and (j)–(l) 18–22 May 1974.
Citation: Journal of Hydrometeorology 12, 4; 10.1175/2010JHM1284.1
(a),(d),(g),(j) SLP (contours every 4 hPa) and IWV (shading every 5 mm) fields. (b),(e),(h),(k) Cross sections of IWV plumes at 80°W showing the cross-barrier component and the tot magnitude of moisture flux (contours and shading every 30 m s1 g kg1, respectively), and (c),(f),(i),(l) time-altitude plots showing winds (barbs; half barb = 5 m s−1 and full barb = 10 m s−1), u wind (full lines every 5 m s1), and IWV (dashed-point line in mm) at 32.5°S–72.5°W grid point of ERA-40 for four heavy precipitation events. The extreme storms of (a)–(c) 5–8 May 1972 and (d)–(f) 24–30 Jun 1974. These storms caused flooding and great damages and casualties in central Chile. The intense storms of (g)–(i) 5–6 Jul 1972 and (j)–(l) 18–22 May 1974.
Citation: Journal of Hydrometeorology 12, 4; 10.1175/2010JHM1284.1
For the remaining events linked with atmospheric rivers, the highest daily precipitation was usually recorded on the day before a cold frontal passage over the crest of the Andes (Fig. 14), which was identified by a drop in the surface pressure, a slight drop in the temperature, an increase in the u wind, and the saturated conditions (i.e., zero dewpoint depression) at the Cristo Redentor site (CRI; selected because it has four-times daily data and is located between the LAG and PIN sites). Furthermore, the 46-case composite mean of daily values at the CRI site, centered on the day of highest precipitation, also suggested that the highest daily precipitation occurs before the cold front passage (i.e., lowest pressure, temperature, and dewpoint depression, and strongest u wind in Table 7), which is consistent with the domain of atmospheric rivers in the pre-cold-front environment of cyclones impinging on the Andes during heavy storms.
Time evolution of the heavy winter storms associated with atmospheric rivers at CRI (on the crest line of the Andes; Table 1) showing the temperature (°C; black solid line with black circles), dewpoint temperature (°C; dotted line with white circles), pressure at the station level (hPa; dark gray line), and u wind (m s−1; gray line with black circles). The gray rectangles on the time axis of each storm panel represent the 24-h periods with highest daily precipitation at LAG–PIN series for each storm respectively.
Citation: Journal of Hydrometeorology 12, 4; 10.1175/2010JHM1284.1
Time evolution of the heavy winter storms associated with atmospheric rivers at CRI (on the crest line of the Andes; Table 1) showing the temperature (°C; black solid line with black circles), dewpoint temperature (°C; dotted line with white circles), pressure at the station level (hPa; dark gray line), and u wind (m s−1; gray line with black circles). The gray rectangles on the time axis of each storm panel represent the 24-h periods with highest daily precipitation at LAG–PIN series for each storm respectively.
Citation: Journal of Hydrometeorology 12, 4; 10.1175/2010JHM1284.1
Time evolution of the heavy winter storms associated with atmospheric rivers at CRI (on the crest line of the Andes; Table 1) showing the temperature (°C; black solid line with black circles), dewpoint temperature (°C; dotted line with white circles), pressure at the station level (hPa; dark gray line), and u wind (m s−1; gray line with black circles). The gray rectangles on the time axis of each storm panel represent the 24-h periods with highest daily precipitation at LAG–PIN series for each storm respectively.
Citation: Journal of Hydrometeorology 12, 4; 10.1175/2010JHM1284.1
Mean daily u wind (m s−1), temperature (°C), dewpoint depression (T − Td, °C), and surface pressure at the CRI station (3800 m; on the crest line of the Andes and between LAG and PIN sites in cross-mountain direction) for one day before highest daily precipitation, the highest precipitation day, and one day after highest daily precipitation at LAG–PIN sites. The mean daily precipitation was calculated from 1200 UTC of the beginning day to 1200 UTC of the following day, agreeing with the date–time of the daily precipitation measurement.
5. Discussion and concluding remarks
The 7-yr 1970–76 climatology of winter precipitation in the SCA (30°–37°S) and the adjacent low areas reveals a strong influence of the mountains on the spatial distribution of precipitation from midlatitude cyclones. The associated synoptic and regional characteristics of this effect are revealed through composites of the 46 heaviest precipitation events (i.e., those that accumulate more than 75% of winter total), subdivided into extreme (95%–100% percentile) and intense (75%–95% percentile) events.
At the large cross-barrier scale (not accounting for the fine ridge–valley scale), the mean winter precipitation field shows peaks on the windward slope below the crest of the Andes, representing an orographic enhancement close to 2 between Portillos, Lagunitas, Sewell, and Chilean low-laying sites around 32.5°–33.5°S—a similar value to that reported by FG07. The precipitation amount drops sharply on the immediate leeward slopes and continues decreasing eastward of the crest, generating a formidable cross-barrier gradient. This large-scale pattern indicates that the SCA are high enough to generate precipitation maximum on the windward slopes below the crest, as was observed in other high-mountain ranges worldwide [e.g., the Cascades and Sierra Nevada in the western United States (Colle and Mass 2000; Steinhauser 1979), the European Alps (Frei and Schär 1998), and the Himalayas (Anders et al. 2006)], rather than on or near the mountain crest, as is the case of lower barriers (e.g., Roe 2005; Anders et al. 2007). Moreover, the winter precipitation field observed in this study is similar to that simulated by VN09 for one heavy case study over the central Andes; this similarity is also consistent with the fact that four or five cool-season heavy events accounted for most of the winter total.
The rain shadow effect is greater in the northern (32°–34.5°S) than in the southern (35°–36.5°S) low-lee side. The difference is probably due to increased baroclinic activity in southernmost latitudes and a southward decrease in the height of the barrier, which allows more precipitation to overflow through the lower Andes. The effect of the SCA range on winter precipitation is so marked that it modifies the precipitation regimes between east and west low adjacent faces north of 35°S. The regime is a winter one in the low-windward side and the windward and immediate leeward slopes, which receives precipitation from storm tracks, while the regime is a summer one in the low-lee side north of 35°S, which receives convective rain. Conversely, the seasonal precipitation has two maxima in the low-lee side between 35° and 37°S—one in winter produced by spillover precipitation and the other in summer produced by convective rain.
The composite synoptic fields reveal that the extreme precipitation events are associated with deeper extratropical cyclones than that for intense events. These deeper cyclones lead to stronger cross-barrier moisture flow against the high Andes that results in heavier orographic precipitation for extreme episodes and, thus, suggests more accentuated upslope moisture flux raining out over the mountains. These results agree with those found by Pandey et al. (1999) and Junker et al. (2008) for winter composite events over the Sierra Nevada. Moreover, the considerable differences in anomaly fields between extreme and intense episodes would have a significant impact on weather forecasting of heavy precipitation in the Andes, which is in agreement with encouraging results in cool-season precipitation events forecast along the U.S. west coast at the Hydrometeorological Prediction Center (Junker et al. 2009).
In spite of the coarse resolution of ERA-40 reanalysis, it represents the narrow IWV plumes associated with a strong water vapor transport toward the Andes in a pre-cold-front environment of cyclones during the heaviest orographic precipitation events, showing stronger IWV plumes and moisture fluxes in extreme than intense events. The individual inspection of each event indicates that 80% of 46 heavy events verified the linkage with the atmospheric river impinging on the Andes just before the cold front passage. The remaining 20% of cases not linked to atmospheric rivers correspond to cut-off lows at middle levels and post-cold-front and warm front synoptic situations (i.e., not pre-cold-front environments). Anomalous low-level northwesterly winds predominated in almost all the events, as exhibited by the rawinsonde data exceeding 2.5 and 2σ in along- and cross-barrier components in 55% and 70% of the 46 cases, respectively. These results coincide with composite studies of FG07 and Neiman et al. (2008), which documented the predominant pre-cold-front environment and strong cross-mountain flux linked to winter precipitation over central Chile and the western coast of North America, respectively. Specifically, Neiman et al. (2008) and Guan et al. (2010) also found that winter storms linked with atmospheric rivers produced twice as much precipitation as the remaining storms. Stronger IWV plumes and IVT resulted in heavier events, which coincides with our results and supports the fact that atmospheric rivers would significantly modulate winter precipitation over the SCA. However, detailed observations over the southeastern Pacific Ocean should be performed to better quantify the concentrated water vapor transport amounts, as documented by Ralph et al. (2004, 2005a) on the California coast.
This study also documented regional airflow characteristics around the Andes, which lead to the observed large-scale orographic precipitation pattern. Mean vertical profiles suggest that the cross-barrier flow would be partially diverted poleward for the high Andes in the form of an along-barrier jet at about 700 hPa, which is still below the crest. Additionally, the typical stable profile at low levels during the heavy storms—likely affected by subtropical Pacific anticyclone—would favor a “flow-around or blocked regime” flow pattern, which is characterized by the upstream blocking and along-barrier jets below the crest of high-mountain ranges [e.g., in the European Alps (Medina and Houze 2003), in the Cascades (Garvet et al. 2007), and in the Sierra Nevada (Marwitz 1987)]. Although observation of water vapor content should be used to confirm the prevailing low-level stability during rainfall times, upstream blocking and along-barrier jets under humid stable conditions have already been observed for heavy case studies in the Andes (Barrett et al. 2009; VN09).
The strong low-level northerly flow could favor the large north–south mean precipitation gradient on the low-windward side through mechanisms proposed by Barrett et al. (2009)—that is, slowing the equatorward advance of the cold front, and increasing the moisture flux convergence along the frontal surface over the southernmost area of central Chile. Nonetheless, upslope flow leading to orographic-enhanced and spillover precipitation on the windward and immediate leeward slopes, respectively, would be restricted to midlevel westerly flow, as suggested by the reduced stability and the cross-barrier wind profile exceeding 10 m s−1 at levels higher than 700 hPa. Previous modeling case studies also support this speculation: VN09 showed that maximum ascent was limited near the upstream crest level, and Barrett et al. (2009) found by trajectory analyses that upstream parcels at low levels did not cross the Andes, whereas parcels at about 700 hPa rose and crossed the Andes. On the low-lee side, the downslope flow produces a prominent rain shadow effect that appears to dominate during heavy events, as denoted by the occurrence of Zonda winds in the majority of cases.
Finally, the typical synoptic and regional conditions associated with heavy orographic precipitation events are summarized in the conceptual representation of Fig. 15. An extratropical cyclone is located on the Pacific coast of South America, with an atmospheric river within the pre-cold-front environment impinging and raining out on the Andes. In addition, the low-level barrier jets upstream and the downslope windstorms downstream of the Andes are regional features typically observed. Future modeling work and higher density data networks are needed to increase our understanding of the highly simplified picture of heavy orographic precipitation over the Andes presented in this study.
Conceptual representation of typical synoptic and regional conditions during the heavy orographic precipitation events over the SCA. The white arrow along the cold front associated with the extratropical cyclone corresponds to strong water vapor transport (i.e., an atmospheric river) toward the Andes, while gray filled arrows correspond to along-barrier jet and downslope flow up- and downstream of the Andes, respectively. Typical weather conditions up- and downstream of the Andes are indicated by rain, snow, orographic clouds, and downslope windstorm symbols.
Citation: Journal of Hydrometeorology 12, 4; 10.1175/2010JHM1284.1
Conceptual representation of typical synoptic and regional conditions during the heavy orographic precipitation events over the SCA. The white arrow along the cold front associated with the extratropical cyclone corresponds to strong water vapor transport (i.e., an atmospheric river) toward the Andes, while gray filled arrows correspond to along-barrier jet and downslope flow up- and downstream of the Andes, respectively. Typical weather conditions up- and downstream of the Andes are indicated by rain, snow, orographic clouds, and downslope windstorm symbols.
Citation: Journal of Hydrometeorology 12, 4; 10.1175/2010JHM1284.1
Conceptual representation of typical synoptic and regional conditions during the heavy orographic precipitation events over the SCA. The white arrow along the cold front associated with the extratropical cyclone corresponds to strong water vapor transport (i.e., an atmospheric river) toward the Andes, while gray filled arrows correspond to along-barrier jet and downslope flow up- and downstream of the Andes, respectively. Typical weather conditions up- and downstream of the Andes are indicated by rain, snow, orographic clouds, and downslope windstorm symbols.
Citation: Journal of Hydrometeorology 12, 4; 10.1175/2010JHM1284.1
Acknowledgments
The authors thank the SMN from Argentina and Dr. Rene Garreaud for providing the daily rainfall series for Argentina and Chile, respectively. We also thank Dr. Mariano Masiokas for providing snow data form high-mountain stations and Dr. Diego Araneo for the computer work involved in the calculations for this study. The authors wish to thank Dr. Gregory Hoke for helping to revise the English in our manuscript. We greatly appreciate the work of observers in situ in the high mountains, especially at the Cristo Redentor site on the crest of the Andes, under extremely bad weather conditions during heavy winter storms. This study was partly funded with Grant CONICET–PIP 112-200801-00195 and UBACYT x160. Finally, the authors would like to thank three anonymous reviewers for their detailed and insightful comments.
REFERENCES
Anders, A. M., Roe G. H. , Hallet B. , Montgomery D. R. , Finnegan N. , and Putkonen J. , 2006: Spatial patterns of precipitation and topography in the Himalaya. Tectonics, Climate, and Landscape Evolution: Geological Society of America Special Paper 398, S. D. Willett et al., Eds., Geological Society of America, 39–53.
Anders, A. M., Roe G. H. , Durran D. R. , and Minder J. R. , 2007: Small-scale spatial gradients in climatological precipitation on the Olympic Peninsula. J. Hydrometeor., 8, 1068–1081.
Barrett, B. S., Garreaud R. D. , and Falvey M. , 2009: Effect of the Andes Cordillera on precipitation from a midlatitude cold front. Mon. Wea. Rev., 137, 3092–3109.
Browning, K. A., 1990: Organization of clouds and precipitation in extratropical cyclones. Extratropical Cyclones: Erik Palmén Memorial Volume, C. W. Newton and E. O. Holopainen, Eds., Amer. Meteor. Soc., 129–153.
Colle, B. A., and Mass C. F. , 2000: The 5–9 February 1996 flooding event over the Pacific Northwest: Sensitivity studies and evaluation of the MM5 precipitation forecasts. Mon. Wea. Rev., 128, 593–617.
Compagnucci, R. H., and Vargas W. M. , 1998: Interannual variability of the Cuyo rivers’ streamflow in the Argentinean Andean mountains and ENSO events. Int. J. Climatol., 18, 1593–1609.
Compagnucci, R. H., and Araneo D. C. , 2005: Identification of areas of statistical homogeneity for the Argentine Andean river flows and its relation to atmospheric circulation and sea surface temperature (in Spanish). Meteorologica, 30, 41–55.
Cressie, N., 1993: Statistics for Spatial Data. John Wiley and Sons, 900 pp.
Durre, I., Vose R. S. , and Wuertz D. B. , 2006: Overview of the Integrated Global Radiosonde Archive. J. Climate, 19, 53–68.
Ereño, C. E., and Hoffman J. , 1978: The Pluvial Regime in the Central Cordillera (in Spanish). Geography Notebook Series, Vol. 5, Faculty of Philosophy and Letters, University of Buenos Aires, 17 pp. [Available from Library of the Faculty of Philosophy and Letters, Institute of Geography “R. Ardissone,” 4th floor, Puan 470, Buenos Aires, Argentina.]
Falvey, M., and Garreaud R. , 2007: Wintertime precipitation episodes in Central Chile: Associated meteorological conditions and orographic influences. J. Hydrometeor., 8, 171–193.
Frei, C., and Schär C. , 1998: A precipitation climatology of the Alps from high-resolution rain gauge observations. Int. J. Climatol., 18, 873–900.
Garvet, M. F., Smull B. , and Mass C. , 2007: Multiscale mountain waves influencing a major orographic precipitation event. J. Atmos. Sci., 64, 711–736.
Goovaerts, P., 2000: Geostatistical approaches for incorporating elevation into the spatial interpolation of rainfall. J. Hydrol., 228, 113–129.
Graham, R. A., and Grumm R. H. , 2010: Utilizing normalized anomalies to assess synoptic-scale weather events in the western United States. Wea. Forecasting, 25, 428–445.
Griffiths, G. A., and McSaveney M. J. , 1983: Distribution of mean annual precipitation across some steepland regions of New Zealand. N. Z. J. Sci., 26, 197–209.
Grumm, R. H., and Hart R. E. , 2001: Standardized anomalies applied to significant cold season weather events: Preliminary findings. Wea. Forecasting, 16, 736–754.
Guan, B., Molotch N. P. , Waliser D. E. , Fetzer E. J. , and Neimann P. J. , 2010: Extreme snowfall events linked to atmospheric rivers and surface air temperature via satellite measurements. Geophys. Res. Lett., 37, L20401, doi:10.1029/2010GL044696.
Hart, R. E., and Grumm R. H. , 2001: Using normalized climatological anomalies to rank synoptic-scale events objectively. Mon. Wea. Rev., 129, 2426–2442.
Hoffman, J. A. J., 1975: Climatic Atlas of South America. Part I: Maps of Mean Temperature and Precipitation. WMO, 4 pp.
Hoskins, B. J., and Hodges K. I. , 2005: A new perspective on Southern Hemisphere storm tracks. J. Climate, 18, 4108–4129.
Junker, N. W., Grum R. H. , Hart R. , Bosart L. F. , Bell K. M. , and Pereira F. J. , 2008: Use of normalized anomaly fields to anticipate extreme rainfall in the mountains of northern California. Wea. Forecasting, 23, 336–356.
Junker, N. W., Brennan M. J. , Pereira F. J. , Bodner M. J. , and Grumm R. H. , 2009: Assessing the potential for rare precipitation events with standardized anomalies and ensemble guidance at the Hydrometeorological Prediction Center. Bull. Amer. Meteor. Soc., 90, 445–453.
Kalnay, E., and Coauthors, 1996: The NCEP/NCAR 40-Year Reanalysis Project. Bull. Amer. Meteor. Soc., 77, 437–471.
Lackmann, G. M., and Gyakum J. R. , 1999: Heavy cold-season precipitation in the northwestern United States: Synoptic climatology and an analysis of the flood of 17–18 January 1986. Wea. Forecasting, 14, 687–700.
Leung, L. R., and Qian Y. , 2003: The sensitivity of precipitation and snowpack simulation to model resolution via nesting in regions of complex terrain. J. Hydrometeor., 4, 1025–1043.
Marwitz, J. D., 1987: Deep orographic storms over the Sierra Nevada. Part I: Thermodynamic and kinematic structure. J. Atmos. Sci., 44, 159–173.
Masiokas, M. H., Villalba R. , Luckman B. H. , Le Quesne C. , and Aravena J. C. , 2006: Snowpack variations in the Central Andes of Argentina and Chile, 1951–2005: Large-scale atmospheric influences and implications for water resources in the region. J. Climate, 19, 6334–6352.
McCauley, M. P., and Sturman A. P. , 1999: A study of orographic blocking and barrier wind development upstream of the Southern Alps, New Zealand. Meteor. Atmos. Phys., 70, 121–131.
Medina, S., and Houze R. A. Jr., 2003: Air motions and precipitation growth for orographic precipitation enhancement. Quart. J. Roy. Meteor. Soc., 129, 345–371.
Miller, A., 1976: The climate of Chile. World Survey of Climatology, W. Schwerdtfeger, Ed., Vol. 12, Elsevier, 113–130.
Minder, J. R., Durran D. R. , Roe G. H. , and Anders A. M. , 2008: The climatology of small-scale orographic precipitation over the Olympic Mountains: Pattern and process. Quart. J. Roy. Meteor. Soc., 134, 817–839.
Montecinos, A., Díaz A. , and Aceituno P. , 2000: Seasonal diagnostic and predictability of rainfall in subtropical South America based on tropical Pacific SST. J. Climate, 13, 746–758.
Neiman, P. J., Ralph F. M. , Wick G. A. , Lundquist J. D. , and Dettinger M. D. , 2008: Meteorological characteristics and overland precipitation impacts of atmospheric rivers affecting the west coast of North America based on eight years of SSM/I satellite observations. J. Hydrometeor., 9, 22–47.
Norte, F. A., 1988: Características del viento zonda en la región de Cuyo–Argentina (Characteristics of the zonda wind in the Cuyo–Argentina region). Ph.D. dissertation, University of Buenos Aires, 255 pp.
Norte, F. A., Ulke A. G. , Simonelli S. C. , and Viale M. , 2008: The severe zonda wind event of 11 July 2006 east of the Andes Cordillera (Argentine): A case study using the BRAMS model. Meteor. Atmos. Phys., 102, 1–14.
Pandey, G. R., Cayan D. R. , and Georgakakos K. P. , 1999: Precipitation structure in the Sierra Nevada of California during winter. J. Geophys. Res., 104 (D10), 12 019–12 030.
Parish, T. R., 1982: Barrier winds along the Sierra Nevada Mountains. J. Appl. Meteor., 21, 925–930.
Philips, D. L., Dolph J. , and Marks D. , 1992: A comparison of geostatistical procedures for spatial analysis of precipitation in mountainous terrain. Agric. For. Meteor., 58, 119–141.
Prohaska, F., 1976: The climate of Argentina, Paraguay, and Uruguay. World Survey of Climatology, W. Schwerdtfeger, Ed., Vol. 12, Elsevier, 13–73.
Ralph, F. M., and Coauthors, 1999: The California Landfalling Jets Experiment (CALJET): Objectives and design of a coastal atmosphere–ocean observing system deployed during a strong El Niño. Preprints, Third Symp. on Integrated Observing Systems, Dallas, TX, Amer. Meteor. Soc., 78–81.
Ralph, F. M., Neiman P. J. , and Wick G. A. , 2004: Satellite and CALJET aircraft observations of atmospheric rivers over the eastern North Pacific Ocean during the winter of 1997/98. Mon. Wea. Rev., 132, 1721–1745.
Ralph, F. M., Neiman P. J. , and Rotunno R. , 2005a: Dropsonde observations in low-level jets over the northeastern Pacific Ocean from CALJET-1998 and PACJET-2001: Mean vertical profile and atmospheric river characteristics. Mon. Wea. Rev., 133, 889–910.
Ralph, F. M., and Coauthors, 2005b: Improving short-term (0–48 h) cool-season quantitative precipitation forecasting: Recommendations from a USWRP workshop. Bull. Amer. Meteor. Soc., 86, 1619–1632.
Ralph, F. M., Neiman P. J. , Wick G. A. , Gutman S. I. , Dettinger M. D. , Cayan D. R. , and White A. B. , 2006: Flooding on California’s Russian River: Role of atmospheric river. Geophys. Res. Lett., 33, L13801, doi:10.1029/2006GL026689.
Ralph, F. M., Sukovich E. , Reynolds D. , Dettinger M. , Weagle S. , Clark W. , and Neiman P. J. , 2010: Assessment of extreme quantitative precipitation forecasts and development of regional extreme event thresholds using data from HMT-2006 and COOP observers. J. Hydrometeor., 11, 1286–1304.
Roe, G. H., 2005: Orographic precipitation. Annu. Rev. Earth Planet. Sci., 33, 645–671.
Satyamurty, P., Nobre C. A. , and Silva Dias P. L. , 1999: South America. Meteorology of the Southern Hemisphere, Meteor. Monogr., No. 49, Amer. Meteor. Soc., 119–139.
Seluchi, M. E., Norte F. A. , Satyamurty P. , and Chou S. C. , 2003: Analysis of three situations of the foehn effect over the Andes (zonda wind) using the Eta-CPTEC regional model. Wea. Forecasting, 18, 481–501.
Sinclair, M. R., Wratt D. S. , Henderson R. D. , and Gray W. R. , 1997: Factors affecting the distribution and spillover of precipitation in the Southern Alps of New Zealand—A case study. J. Appl. Meteor., 36, 428–442.
Smith, R. B., and Evans J. P. , 2007: Orographic precipitation and isotope fraction over the southern Andes. J. Hydrometeor., 8, 3–14.
Smith, R. B., Barstad I. , and Bonneau L. , 2005: Orographic precipitation and Oregon’s climate transition. J. Atmos. Sci., 62, 177–191.
Steinhauser, F., 1979: Climatic Atlas of North and Center America. WMO, 4 pp.
Stuart, N. A., and Grumm R. H. , 2006: Using wind anomalies to forecast East Coast winter storms. Wea. Forecasting, 21, 952–968.
Trenberth, K. E., 1991: Storm tracks in the Southern Hemisphere. J. Atmos. Sci., 48, 2159–2178.
Uppala, S. M., and Coauthors, 2005: The ERA-40 Re-Analysis. Quart. J. Roy. Meteor. Soc., 131, 2961–3012.
Viale, M., 2010: Characteristics of winter orographic precipitation over the subtropical central Andes (in Spanish). Ph.D. dissertation, University of Buenos Aires, 163 pp. [Available online at http://digital.bl.fcen.uba.ar/Download/Tesis/Tesis_4707_Viale.pdf.]
Viale, M., and Norte F. A. , 2009: Strong cross-barrier flow under stable conditions producing intense winter orographic precipitation: A case study over the subtropical central Andes. Wea. Forecasting, 24, 1009–1031.
Wratt, D. S., Revell M. J. , Sinclair M. R. , Gray W. R. , Henderson R. D. , and Chater A. M. , 2000: Relationships between air mass properties and mesoscale rainfall in New Zealand’s Southern Alps. Atmos. Res., 52, 261–282.
Zhu, Y., and Newell R. E. , 1998: A proposed algorithm for moisture fluxes from atmospheric rivers. Mon. Wea. Rev., 126, 725–735.
