The El Niño Southern Oscillation (ENSO) phenomenon is the major single source of seasonal-to-interannual climate variability (Grimm et al. 2000; Montecinos et al. 2000). There are marked links between ENSO and precipitation in the Pampas in Nov-Dec, a critical period for important summer crops. In these months, El Niño events are associated with higher median precipitation and higher likelihood of positive (wet) rainfall anomalies than other ENSO phases, whereas La Niña events show markedly lower median rainfall and a narrower range of anomalies (Podestá et al. 1999; Rusticucci & Vargas 2002).

Precipitation data at 23 locations (Figure 1) for the period 1912­1990 were aggregated into four quarterly series of precipitation totals. The quarterly series were July-September (as we defined the “ENSO year” to begin in July, Podestá et al., 1999) through April–June. The data come from cooperative observing stations (i.e., not meteorological stations) but they have been quality controlled. Furthermore, the PCA approach drastically reduces effects of possible inhomogeneities or erroneous values in the original data series. For each quarter, the temporal mean precipitation total for each location was subtracted from the corresponding precipitation series. PCA decompositions were performed on the correlation matrix of the resulting quarterly time series of precipitation anomalies. 

The time series of the first principal component (PC1, also referred to as amplitudes or scores) can be viewed as an optimally weighted average (where the weights are estimated through the PCA) of precipitation anomalies throughout the area. Therefore, the PC1 scores summarize in a single series the temporal evolution of precipitation anomalies over the Pampas for a given quarter. The percentage of variance accounted for by the PC1 was 37, 44, 36, and 47 for JAS, OND, JFM, and AMJ, respectively. The summary series for each quarter subsequently were analyzed for association with ENSO phase. Although other interesting aspects could be derived from the PCA results (spatial homogeneity of precipitation variability, low-frequency variability), we focused on the associations with ENSO, as the ultimate goal was to link the Pampas-wide precipitation anomalies with crop yields at national level.

An ENSO phase was assigned to each record in the four series of quarterly precipitation anomalies. ENSO phase was defined based on sea surface temperature anomalies in the tropical Pacific Ocean between 4°N–4°S and 90°W–150°W (see Podestá et al., 1999 for details). Boxplots of PC1 amplitudes by ENSO phase are shown for each quarter in Figure 2. Although PC amplitudes have no physical units and their signs are arbitrary, in this case positive PC1 values indicate positive precipitation anomalies, and vice-versa.

There is a distinct association between ENSO phase and precipitation in the Pampas during OND (Figure 2b). Warm ENSO events (or Niño years) are associated with slightly higher median precipitation and a higher proportion of high anomalies than other phases, whereas cold (Niña) events show markedly lower median precipitation and a much smaller range of anomalies. ENSO signals during other quarters are not as apparent. The Pampas-wide patterns summarized through PCA can be illustrated with specific precipitation values at certain locations. For example, in Pergamino, one of the top-producing agricultural locations, OND median total precipitation for warm, neutral, and cold events is 349, 332, and 185 mm. Pairwise Wilcoxon tests for this location showed significant differences in the central tendency of precipitation between cold and neutral events (P << 0.001), but the values were not significantly different for warm and neutral events (P = 0.624).

Return to General Background page.

In other NSF-suported work we linked climatic, agronomic, and financial models to characterize vulnerability to ENSO of current maize production systems in Argentina (Ferreyra et al. 2001). We combined synthetic weather conditioned on ENSO phase (Grondona et al. 2000), a maize simulation model (Ritchie et al. 1998) and stochastic prices in an enterprise budget to derive probability distributions of profits for each ENSO phase. Strong consistency between model and historical results suggests this approach can be used confidently in this project.

The goal of enhancing and sustaining the ability of decision-makers to use climate information can be accomplished most effectively through existing “boundary organizations” that perform information communication and translation (Guston et al. 2000; Agrawala et al. 2001; Cash et al. 2003).

Our partner Asociación Argentina de Consorcios Regionales de Experimentación Agrícola (AACREA) is an example of a boundary organization. AACREA is a non-governmental, non-profit organization of farmers with a focus on dissemination of new technologies. The main characteristic of AACREA is teamwork.

AACREA members join groups of 7–12 farmers. Each group has a technical advisor (funded by group members) who provides information and advice to group members and coordinates exchanges among the group. Each group meets monthly, a ready-made opportunity for project researchers to interact with group members. Currently, about 1300 farmers are members of 140 CREA groups in 17 different ecological regions of Argentina. AACREA members dedicate about 2.5 Mha to agricultural and cattle production; they contribute from 6% to 20% of total production (depending on the specific activity considered) in Argentina.

AACREA is an excellent example of a boundary organization. Although the organization does not conduct its own research, AACREA’s national headquarters employ scientists who commission research from universities or other institutions and adapt research results to the needs of the farmer members. At the same time, a major feature of AACREA is its strong commitment towards dissemination of technological innovations. The dissemination takes place (a) among members of a group during regular monthly meetings, (b) at “open farmgate” meetings open to non-members, (c) at regional, national and international meetings, and (d) through AACREA’s monthly magazine and numerous technical publications. As a result, it has been estimated that AACREA has a considerable multiplying effect on agricultural technology dissemination: for each AACREA member, information reaches about 40 other farmers (i.e., about 52,000 farmers in Argentina). AACREA has a leading role in agricultural technology dissemination because budgetary problems have weakened significantly the governmental agricultural extension system.

The Argentine Meteorological Service (Servicio Meteorológico Nacional, SMN) is a governmental organization charged with collecting, analyzing, and disseminating weather and climate information. During focus groups we conducted in the Pampas during a previous Biocomplexity incubation effort, participating farmers identified SMN as their primary source of climate information (according to an index that weighted the number of times SMN was mentioned by the order in which it was listed relative to other possible sources).

SMN also can be considered a boundary organization, as it has a dual mandate for producing and communicating climate information. SMN’s Climatology Division routinely produces a monthly climate outlook and represents the institution in regional climate fora periodically held in southeastern South America (Buizer et al., 2000). The Agroclimatology Division regularly produces diagnostic information useful to farmers, such as precipitation anomaly maps, soil water content, and satellite-derived vegetation indices. Nevertheless, SMN’s linkages to agricultural stakeholders are much less direct than AACREA’s and, to our knowledge, formal mechanisms to receive feedback and respond to it are virtually inexistent.

The use of climate information is rapidly evolving from decisions based on analysis of historical records to an increasing capability to monitor and predict seasonal regional climate. The increase in scientific and technological capabilities coincides with a growing appreciation of the importance of climate in human endeavors, leading to a large increase in the demand for climate information. It is likely that in the near future SMN will be expected to play a role as a “climate service” (National Research Council, 2001), in addition to its traditional function of monitoring and predicting weather. However, it is unclear if this institution (like many of its sister meteorological and hydrological agencies elsewhere) is equipped to fulfill this role.

Return to General Background page.