Abstract. As part of the study on the Mantaro river basin's (central Andes of Perú) current vulnerability to climate change, the temporal and spatial characteristics of frosts were analysed. These characteristics included intensity, frequency, duration, frost-free periods, area distribution and historical trends. Maps of frost risk were determined for the entire river basin, by means of mathematical algorithms and GIS (Geographic Information Systems) tools, using minimum temperature – 1960 to 2002 period, geomorphology, slope, land-use, types of soils, vegetation and life zones, emphasizing the rainy season (September to April), when the impacts of frost on agriculture are most severe. We recognized four categories of frost risks: low, moderate, high and critical. The critical risks (with a very high probability of occurrence) were related to high altitudes on the basin (altitudes higher than 3800 m a.s.l.), while the low (or null) probability of occurring risks were found in the lower zones (less than 2500 m a.s.l.). Because of the very intense agricultural activity and the high sensitivity of the main crops (Maize, potato, artichoke) in the Mantaro valley (altitudes between 3100 and 3300 m a.s.l.), moderate to high frost risks can be expected, with a low to moderate probability of occurrence. Another significant result was a positive trend of 8 days per decade in the number of frost days during the rainy season.

Abstract. A local integrated assessment of the vulnerability and adaptation to climate change in the Mantaro River Basin, located in Peruvian Central Andes, was developed between years 2003 to 2005. In this paper we present some lessons learned during the development of this study, emphasizing the multi-institutional and interdisciplinary efforts, briefly showing the methodological aspects, and pointing out the main problems found.

The climatological and large‐scale characteristics of the extreme cold events (ECEs) in the central Peruvian Andes (Mantaro basin (MB)) during austral summer (January–March) are examined using reanalysis, gridded and in situ surface minimum temperature (Tmin) data for the 1979–2010 period. To describe the influence of the Madden–Julian Oscillation (MJO) on ECEs in the MB, two ECE groups are defined on the basis of the sign of the outgoing long‐wave radiation (OLR) anomalies in the MJO band (30–100 days, 0–9 eastward) at 12.5°S, 75°W. Type‐1 ECEs occur during the suppressed convection phase of the MJO (OLR anomalies ≥+2 W/m2) while Type‐2 ECEs occur during the enhanced convection phase of the MJO (OLR anomalies ≤−2 W/m2). ECEs in the MB are associated with the advection of cold and dry air along the east of the Andes through equatorward propagation of extratropical Rossby wave trains (ERWTs). This cold advection weakens the Bolivian High–Nordeste Low (BH‐NL) system over South America (SA) at upper‐tropospheric levels. The MJO is an important driver of ECEs in the MB, favouring the cold advection along the Andes during specific MJO phases. Fifty‐nine per cent of Type‐1 ECE's and 86% of Type‐2 ECE's occur in MJO Phases 7‐2. Type‐1 and 2 ECEs feature a weakened BH over SA at upper‐tropospheric levels. For Type‐1, ERWTs emanate from southeastern Africa in MJO Phases 8‐1 while ERWTs are strengthened when crossing the subtropical southern Pacific Ocean during MJO Phases 2 and 7. With respect to Type‐2, MJO Phases 7‐2 feature circumpolar Rossby wave trains propagating toward SA. Ultimately, MJO Phases 7‐2 induce negative Tmin anomalies over MB, while MJO Phases 3‐6 induce positive Tmin anomalies. El Niño and La Niña strengthen negative Tmin anomalies over the MB during MJO Phases 7‐8 while they weaken positive Tmin anomalies over the MB during MJO Phases 3‐6.

The most extreme precipitation event in Metropolitan Lima (ML) occurred on 15 January 1970 (16 mm), this event caused serious damage, and the real vulnerability of this city was evidenced; the population is still not prepared to resist events of this nature. This research describes the local climate variability and extreme climate indices of temperature and precipitation. In addition, the most extreme precipitation event in ML is analyzed. Extreme climate indices were identified based on the methodology proposed by the Expert Team on Climate Change Detection and Indices (ETCCDI). Some extreme temperature indices highlight an initial trend toward warm conditions (1965–1998); this trend has changed towards cold conditions since 1999, consistent with the thermal cooling during the last two decades in ML (−0.5 °C/decade) and other coastal areas of Peru. The variations of extreme temperature indices are mainly modulated by sea-surface temperature (SST) alterations in the Niño 1 + 2 region (moderate to strong correlations were found). Extreme precipitation indices show trends toward wet conditions after the 1980s, the influence of the Pacific Ocean SST on the extreme precipitation indices in ML is weak and variable in sign. The most extreme precipitation event in ML is associated with a convergence process between moisture fluxes from the east (Amazon region) at high and mid levels and moisture fluxes from the west (Pacific Ocean) at low levels, and near the surface.