Abstract

The present research has assessed the spatial distribution of drought risk in the Titicaca Lake Basin located in Peruvian territory for a district scale, based on hazard and vulnerability. Drought hazard has been quantified with the deficit of precipitation using the Standardized Precipitation Index (SPI) for a time scale of 3-months, and the vulnerability has been obtained according to the socio-economic and physical indicators of the Basin. The results show that the drought is strongly modulated by anomalous SST conditions of the surrounding Oceans, mainly by the Pacific region. Over the Titicaca Lake Basin was identified that about 50% of districts present a high to very high risk of drought mainly, in the northwestern, central-east, and central-south of the Basin. These districts have a larger deficit of precipitation and showed indicators that are more vulnerable to the drought hazard due to that economically depends on precarious rainfed agriculture and an extensive and mixed livestock system. Within this area are the most important provinces of the high-Andean region, such as Puno, San Roman, Azángaro, Melgar, and Carabaya, and outstanding districts as Puno and Juliaca, considered the economic capital of the department of Puno.

Keywords
Titicaca Lake basin; drought; risk; hazard; vulnerability

Resumo

A presente pesquisa avaliou a distribuição espacial do risco de seca na bacia hidrográfica do lago Titicaca localizada em território peruano para uma escala de distrito, com base em perigo e vulnerabilidade. O perigo foi quantificado com o déficit de precipitação usando índice de Precipitação Padronizado (SPI, por suas siglas em inglês), para uma escala de tempo de três meses; e a vulnerabilidade foi obtida de acordo com os indicadores socioeconômicas e físicas da bacia. Os resultados mostram que a seca é fortemente modulada por condições anômalas de TSM dos oceanos circundantes, principalmente pela região do Pacífico. Na Bacia do Lago Titicaca, foi identificado que aproximadamente 50% dos distritos apresentam alto a muito alto risco de seca, localizados principalmente no Noroeste, centro-leste, e centro-sul da bacia. Esses distritos apresentam um maior déficit de precipitação e apresentam indicadores mais vulneráveis ao perigo de seca devido ao facto de depender economicamente de uma agricultura de sequeiro precária e de um sistema de pecuária extensiva e mista. Nesta área estão as províncias mais importantes da região alto-andino, como Puno, San Roman, Azángaro, Melgar e Carabaya, e destacam-se distritos como Puno e Juliaca considerado o capital econômico do departamento de Puno.

Palavras-chave
Bacia hidrográfica do lago Titicaca; seca; risco; perigo; vulnerabilidade

1. Introduction

Drought is one of the most damaging natural hazards related to climate (Sönmez et al., 2005SöNMEZ, K.; KöMüSCü, A.; ERKAN, A.; TURGU, E. An analysis of spatial and temporal dimension of drought vulnerability in Turkey using the standardized precipitation index. Natural Hazards, v. 35, n. 2, p. 243-264, 2005.), and is considered to affect more people than any other natural disaster on the planet (Ortega, 2013ORTEGA, D. Sequía: Causas y efectos de un fenómeno global. Ciencia UANL, n. 61, p. 9, 2013.). It is therefore considered by many to be the most complex but least understood of all-natural hazards (WMO, 2016WMO. Handbook of Drought Indicators and Indices. Integrated Drought Management Programme (IDMP), Integrated Drought Management Tools and Guidelines Series 2. Geneva: WMO, p. 52, 2016.). It differs from other hazards such as floods, tropical cyclones, and earthquakes because it builds up slowly over a considerable period and can persist for years after the event has ended, making it difficult to determine the beginning and end (Tannehill, 1947TANNEHILL, I. Drought, its causes and effects, Princeton, NJ: Princeton University Press. American Journal of Agricultural Economics, v. 29, n. 3, p. 779-781, 1947.).

Usually, droughts are classified according to type, in meteorological, agricultural, and hydrological, and they are differentiated by their intensity, duration, and spatial extent (OMM, 2012OMM. Guía del usuario sobre el índice normalizado de precipitación (OMM-Nº 1090). Geneva: Organización Meteorológica Mundial, p. 23, 2012.). Meteorological drought is defined as a threshold of precipitation deficit that is reached during a period previously determined (OMM, 2006OMM. Vigilancia y Alerta Temprana de la Sequía: Conceptos, Progresos y Desafíos Futuros (OMM-Nº 1006). Geneva: Organización Meteorológica Mundial, p. 28, 2006.). The majority of available drought-related indices focus on the assessment of meteorological drought, for example, Standard Precipitation Index (SPI), Palmer Index (PI), Simple Precipitation Intensity Index (SPII) (Rojas, 2021ROJAS, O. Next Generation Agricultural Stress Index System (ASIS) for agricultural drought monitoring. Remote Sens, v. 13, n. 5, p. 959, 2021.). The index most used is the Standardized Precipitation Index (SPI), which is recommended by the World Meteorological Organization (WMO) as the standard drought index (Serafini et al., 2019SERAFINI, B.; OLIVEIRA-JúNIOR, J.; DE GOIS, G.; PEREIRA-JúNIOR, E. Drought characterization for the state of Rio de Janeiro based on the annual SPI index: trends, statistical tests and its relation with ENSO. Atmospheric Research, v. 220, p. 141-154, 2019.) given its simplicity and flexibility for the study of precipitation on various time scales (Endara et al., 2019ENDARA, S.; ACUñA, J.; VEGA, F.; FEBRE, C.; CORREA, K; et al. Caracterización Espacio Temporal de la Sequía en los Departamentos Altoandinos del Perú (1981-2018). Lima: Servicio Nacional de Meteorología e Hidrología del Perú, p. 31, 2019.). Mckee et al. (1993)MCKEE, T.; DOESKEN, N.; KLEIST, J. The Relationship of drought frequency and duration to time scales. In: Eighth Conference on Applied Climatology, Anaheim, p. 17-22, 1993. originally developed the SPI to quantify the precipitation deficit, for different time scales of 3, 6, 12, 24, and 48 months. Short-term scales such as months may be important for agricultural interests, while long-term scales such as years may be transcendental for interests in water supply management (Guttman, 1998GUTTMAN, N. Comparing the palmer drought index and the standardized precipitation index. Journal of the American Water Resources Association, v. 34, n. 1, p. 113-121,1998.).

In Peru, the droughts are quite often since last decades, according to Lovón (1985)LOVóN, G. El Sur Andino Peruano y la Coyuntura de Sequía: 1982-1983. Buenos Aires: Grupo Editor Latinoamericano, p. 17, 1985.. Between 2000 and 2010, 163 drought events with different frequencies were reported at the national level, being higher on the Basin of Pacific, followed by the Titicaca Lake Basin (ANA, 2013ANA. Las Condiciones de Sequía y Estrategias de Gestión en el Perú. Informe Nacional del Perú. Lima: Autoridad Nacional del Agua, p. 11, 2013.). The Titicaca Lake Basin is located in the southern region of Peru, an area where several drought events have occurred (Vega, 2016VEGA, F. Análisis del Riesgo de Sequias en el Sur del Perú. Lima: Servicio Nacional de Meteorología e Hidrología del Perú, 2016.). According to the National Institute of Civil Defense (INDECI, by acronym in Spanish), historically the area of the southern highlands of Peru is the area most prone to the occurrence of droughts, this area includes the departments of Puno, Cusco, Tacna, Moquegua, Arequipa and Apurimac, with 70% of the population economically active in rainfed agriculture and livestock. One of the most critical occurred at the end of 1982 and the beginning of 1983, it was an exceptional drought that fundamentally affected the highland the Puno and Cusco departments (Lovón, 1985LOVóN, G. El Sur Andino Peruano y la Coyuntura de Sequía: 1982-1983. Buenos Aires: Grupo Editor Latinoamericano, p. 17, 1985.).

The Lake Titicaca Basin region features several characteristics that influence the spatial-temporal rainfall distribution, and many studies have shown that regions with lakes (e.g., Lake Titicaca Basin) are affected by a combination of orographic and convective precipitation (Chuchón and Pereira, 2022CHUCHóN, E.; PEREIRA, A. The diurnal cycle of precipitation over Lake Titicaca Basin based on CMORPH. Atmosphère, v. 13, n. 4, p. 601, 2022.). The rainy season over the Altiplano occurs during the austral summer linked to the active phase of the South America Monsoon System-SAMS (Zhou and Lau 1998ZHOU, J.; LAU, K.M. Does a Monsoon climate exist over South America? Journal of Climate, v. 11, n. 5, p. 1020-1040, 1998.; Vera et al., 2006VERA, C.; HIGGINS, W.; AMADOR, J.; AMBRIZZI, T.; GARREAUD, R.; et al. Toward a unifed view of the American monsoon systems. Journal of Climate, v. 19, n. 20, p. 4977-5000, 2006.). The peak of precipitation occurs in January, and is characterized by intense convective activity combined with moisture advection from the Amazon Basin (Aceituno and Montecinos 1993ACEITUNO, P.; MONTECINOS, A. Circulation anomalies associated with dry and wet periods in the South American Altiplano. In: Proceedings of the Fourth International Conference on Southern Hemisphere Meteorology, Hobart: American Meteorological Society, p. 330-331, 1993.; Chaffaut et al., 1998CHAFFAUT, I.; POUYAUD, B.; MICHELOT, J.L.; COUDRAIN-RIBSTEIN, A. Précipitations d'altitude du Nord-Chili, origine des sources de vapeur et données isotopiques. Bulletin de l'Institut Française d'Etudes Andines, v. 27, n. 3, p. 367-384, 1998.; Vuille et al., 1998VUILLE, M.; DOUGLAS, H.; CARSTEN, F.; RAYMOND, B. Atmospheric circulation anomalies associated with 1996/1997 summer precipitation events on Sajama Ice Cap, Bolivia. Journal of Geophysical Research-Atmospheres, v. 103, n. 10, p. 11191-11204, 1998.; Garreaud et al., 2003GARREAUD, R.; VUILLE, M.; CLEMENT, A. The climate of the Altiplano: Observed current conditions and mechanisms of past changes. Palaeogeography, Palaeoclimatology, Palaeoecology, v. 194, p. 5-22, 2003.; Vizy and Cook, 2007VIZY, E.; COOK, K. Relationship between Amazon and high Andes rainfall. Journal of Geophysical Research-Atmospheres, v. 112, n. 7, p. 14, 2007.; Mayta et al., 2019MAYTA, V.; AMBRIZZI, T.; ESPINOZA, J.C.; SILVA, P. The role of the Madden-Julian oscillation on the Amazon Basin intraseasonal rainfall variability. International Journal of Climatology, v. 39, n. 1, p. 343-360, 2019.). During the austral winter (in the absence of the SAMS influence), the moist eastern flux is replaced by westerly winds, which provide dry air related to the atmospheric stability over the Pacific Ocean. This strong seasonality explains that almost all precipitation occurs in 1-3 month (Lavado et al., 2012LAVADO, W.; JOSYANE, D.; ESPINOZA, J.C.; LOUP, J. Basin-scale analysis of rainfall and runoff in Peru (1969-2004): Pacific, Titicaca and Amazonas drainages. Hydrological Sciences Journal, v. 57, n. 4, p. 625-642, 2012.).

The present research assesses the spatial distribution of drought risk in the Titicaca Lake Basin located in Peruvian territory for a district territorial division, based on hazard and vulnerability. The drought hazard is estimated with the precipitation deficit monthly rainfall obtained from the latest version of product PISCO - Precipitation v1.2 (Peruvian Interpolation of the SENAMHIs Climatological and Hydrological Stations) for the period 1981 to 2016 (36 years), and then is quantified with the Standardized Precipitation Index (SPI) for a time scale of 3-months. The vulnerability is obtained according to socio-economic and physical indicators of the hydrographic Basin of Lake Titicaca.

1.1. Geographical and climatic characteristics of the Titicaca Lake basin

The hydrographic Basin of Lake Titicaca is located in South America, south of Peru and northwest of Bolivia, between 14°05' and 16°50' north longitude and 68°10' and 71°05' west longitude, on the Peruvian territory covers an area of 48,910 km² including the part corresponding to the Lake, with an altitude varying between 3,812 and 5,500 m.a.s.l. In addition, has a cold and dry climate, with a maximum temperature of 18 °C, a minimum of -2 °C, and an average of 10 °C. The spatial distribution of annual mean precipitation has a decreasing pattern from north to south. In general, it varies from 200 to 1,400 mm (INEI, 2013INEI. IV Censo Nacional Agropecuario 2012: PUNO Perfil Agropecuario. Punto: Grafía S.A.C., p. 201, 2013.).

In Peruvian territory, the Titicaca Lake Basin is located within the high-Andean region of the department of Puno, occupying approximately 63% of its surface. This is composed of 13 provinces and 90 districts, as can be seen in Fig. 1 and listed in Table 1. The district with the largest area is Santa Rosa district with 2,472 km², belonging to the province of El Collao, followed by Nuñoa district with 2,209 km², located in the province of Melgar. Likewise, the Titicaca Lake Basin is made up of 16 hydrographic units, as can be seen in Fig. 2. The hydrographic unit of Azángaro has the largest area with 8,754 km², followed by the Pucará with 5,541 km².

2. Data and Methodology

2.1. Data

2.1.1. Precipitation data

Monthly rainfall is obtained from product PISCO-Precipitation v1.2 (Peruvian Interpolation of the SENAMHIs Climatological and Hydrological Stations) for the period 1981 to 2016 (36 years). PISCO - precipitation is a spatial precipitation database produced and validated by the National Meteorological and Hydrological Service of Peru (SENAMHI, by acronym in Spanish) at a monthly time step and a grid resolution of 0.05°. The series starts in January 1981 until 2016 (Aybar et al., 2017AYBAR, C.; LAVADO, W.; HUERTA, A.; FERNáNDEZ, C.; VEGA, F.; et al. Uso del Producto Grillado PISCO de Precipitaciones en Estudios, Investigaciones y Sistemas Operacionales de Monitoreo y Pronostico Hidrometeorológico. Lima: Servicio Nacional de Meteorología e Hidrología del Perú, p. 22, 2017.) and can be freely downloaded through the SENAMHI website. The validation of PISCO-precipitation in the hydrographic regions of Peru is detailed by Endara (2017)ENDARA, S. Determinación de Extremos a Precipitación a Partir del PISCO Diario. Lima: Servicio Nacional de Meteorología e Hidrología del Perú, 83 p., 2017. and Porto (2021)PORTO, A. Validación del Producto Grillado Pisco de Precipitación v2.1 con la Data Observada de las Estaciones Pluviométricas del SENAMHI en la Cuenca de la Vertiente del Lago Titicaca - Lado Peruano. Puno: Universidad Nacional del Altiplano, 2021..

2.1.2. Oceanic indices

To assess the impacts of El Niño-Southern Oscillation (ENSO) in precipitation, we used the eastern (E) and central (C) index which represent anomalous surface warming in the east and center Pacific, as can be seen in Fig. 3 documented in detail by Takahashi et al. (2011)TAKAHASHI, K.; MONTECINOS, A.; GOUBANOVA, K.; DEWITTE, B. ENSO regimes: Reinterpreting the Canonical and Modoki El Niño. Geophysical Research Letters, v. 38, n. 10, p. 5, 2011. and is available at the Climate Monitoring Indices of the Geophysical Institute of Peru. In addition, to measure the impacts of the anomalous condition over Tropical North Atlantic (TNA), we used the TNA index. This index is computed as the anomaly of the average of the monthly SST over 55° W-15° W, 5° N-25° N region and is available from NOAA Climate Indices.

2.1.3. Socioeconomic and physical indicators

The indicators were quantified using the latest National Censuses 2017: Population XII, Housing VII, and Indigenous Communities III; National Agricultural Census IV - 2012. Additionally, the Census population, statistical information of Economically Active Population, and Human Development Index are obtained from websites and can be freely downloaded.

2.2. Methodology

2.2.1. Precipitation analysis

The rainy months are identified for drought monitoring in the Titicaca Lake Basin due to that the main economic activities in the Basin depend on rainfall for their development. The fluctuation of the rainy months and their slope-index are calculated to know if increase or decrease the precipitation trends, and detect if they present statistical significance. The trend analysis was carried out with the statistical test non-parametric Mann-Kendall at a 95% of confidence level.

2.2.2. Drought hazard calculation

The precipitation deficit of meteorological drought in the Titicaca Lake Basin is quantified using the Standardized Precipitation Index (SPI) for a temporary scale of 3-months. The short time scales (1, 2, or 3 months), can provide an early warning of drought and help assess drought severity (OMM, 2012OMM. Guía del usuario sobre el índice normalizado de precipitación (OMM-Nº 1090). Geneva: Organización Meteorológica Mundial, p. 23, 2012.). In the Titicaca Lake Basin, it can be essential because the agricultural activity is the primary source of economic income in the region and it depends on the precipitation for its development. Table 2 shows the category of droughts defined by SPI values. There is evidence that the high-Andean region of Puno presents a strong signal associated with anomalous conditions observed in the Pacific region (i.e., El Niño). Therefore, the correlation between the Peruvian precipitation (from PISCO for January to March season and the 1980-2016 period) against the TNA index, E index, and C index is analyzed.

For the calculation of drought hazard, each district is assigned a weight and a rating using the analytical hierarchy process (AHP), according to the drought category (moderate, severe, or extreme) that depends on the values obtained from the SPI and the percentage of drought occurrence. AHP is a decision-making method that systematically evaluates alternatives to derive priorities and importance when evaluation criteria are numerous and complex (Young-Sik et al., 2020YOUNG-SIK, M.; WON-HO, N.; MIN-GI, J.; NA-KYOUNG, B.; TAEGON, K. Assessment of vulnerability to drought disaster in agricultural reservoirs in South Korea. Journal Atmosphere, v. 11, n. 11, p. 1244, 2020.). The percentage of drought occurrence is calculated by dividing the number of specific drought events (moderate, severe, or extreme) by the total events of the analyzed period. Table 3 shows the weights and qualifications assigned according to drought category and percentage of drought occurrence.

The drought occurrence and hazard maps are developed for a time scale of 3-months at the district scale. The maps are made using the kriging interpolation method which is widely recognized as a standard approach for interpolating surfaces based on scalar measurements of different points (Shahid and Behrawan, 2008). The Drought Hazard Index (DHI) was derived using Eq. (1) according to Hungsoo et al. (2015)HUNGGSOO, K.; JONGYONG, P.; JIYOUNG, Y.; TAE-WOONG, K. Assessment of drought hazard, vulnerability, and risk: A case study for administrative districts in South Korea. Journal of Hydro-environment Research, v. 9, n. 1, p. 28-35, 2015..

where MDw is the weight assigned to moderate drought; MDr is rating assigned to the moderate drought percentage of drought occurrence; SDw is the weight assigned to severe drought; SDr is rating assigned to the severe drought percentage of drought occurrence; VSDw is the weight assigned to extreme drought; VSDr is rating assigned to the extreme drought percentage of drought occurrence.

2.2.3. Vulnerability to drought calculation

The vulnerability to drought is obtained by Eq. (2), which is based on 4-socioeconomic and 2-physical indicators of the Titicaca Lake Basin. The Titicaca Lake Basin is located within the high-Andean region of the department of Puno, which is one of the departments with more poverty at the national level, with main economic activities that depend on the rains for their development such as rainfed agriculture and livestock. For this reason, the socioeconomic and physical indicators are chosen according to the most relevant activities and characteristics of the region which are negatively affected due to the impact of drought events. Table 4 shows the socioeconomic and physical indicators detailed.

The socioeconomic and physical indicators are divided into 4 classes: low, moderate, high, and very high, using the natural breaks method (Jenks). The natural break method is used to derive the classes. This method creates a range according to an algorithm that uses the average of each range to distribute the data more evenly across the ranges (Shahid and Behrawan, 2008SMITH, R. Comparing traditional methods for selecting class intervals on choropleth maps. The Professional Geographer, v. 38, n. 1, p. 62-67, 1986.). This method ensures that the ranges are well represented by their averages and that the data values within each range are fairly close together (Smith, 1986SMITH, R. Comparing traditional methods for selecting class intervals on choropleth maps. The Professional Geographer, v. 38, n. 1, p. 62-67, 1986.). GIS software ArcView 9.0 is used to identify the natural breakpoints in the data.

The vulnerability to drought is obtained with the index that is the average of the scores assigned to each indicator as shown in the Eq. (2) according to Johnson et al. (2016)JOHNSON, K.; DEPIETRI, Y.; BREIL, M. Multi-hazard risk assessment of two Hong Kong districts. International Journal of Disaster Risk Reduction, v. 19, p. 311-323, 2016. doi
doi... . The vulnerability index is classified into four ranges with the Natural Break method (Jenks), according to their 25%, 50% and 75% percentiles, then scores of 1.0, 2.0, 3.0 and 4.0 are placed to distinguish their levels of vulnerability (low, moderate, high and very high). Finally, the drought vulnerability map is prepared at the district level differentiated the vulnerability levels with colors.

where Dp is the score assigned to population density; Id is the score assigned to the Human Development Index; Ae is the score assigned to Active Economic Activity in agriculture, livestock, hunting, and forestry; Rhm is the score assigned to the female/male relation; Pg is the score assigned to the livestock population; Sa is the score assigned to the rainfed agricultural area.

2.2.4. Drought risk calculation

Drought risk is obtained through the Drought Risk Index (DRI) which is a function of hazard and vulnerability, as shown by Eq. (3) according to Shahid and Behrawan (2008)SHAHID, S.; BEHRAWAN, H. Drought risk assessment in the western part of Bangladesh. Natural Hazards, v. 46, n. 3, p. 391-413, 2008..

where DRI is the drought risk index; DHI is the drought hazard index; DVI is the drought vulnerability index. Figure 4 shows the schematic depiction of the methodology used in this study to calculate the drought risk.

3. Results and Discussion

3.1. Precipitation analysis

Figure 5 shows the spatial distribution of the annual mean precipitation (1981-2016) in the hydrographic units of the Titicaca Lake Basin and the rainfall for the January to March period. The precipitation over the Titicaca plateau is important to study because the development of the main industries in this region depends on it (Romero, 2013ROMERO, E. Monografía del Departamento de Puno. 3. ed. Puno: Universidad Nacional del Altiplano, p. 601, 2013.). The hydrographic units with the highest rainfall are located in the west, center, and east of the Titicaca Lake Basin, while those in the north and south show the lowest rainfall. The rainy months are January, February, and March (hereafter, JFM) due to during this period the precipitation is higher compared to the rest of the year, representing more than 50% of the annual accumulated precipitation. July is the driest month, while January is the wettest month (Andrade, 2017ANDRADE, M. Atlas-Clima y eventos extremos del Altiplano Central Perú-Boliviano/Climate and extreme events from the central Altiplano of Peru and Bolivia 1981-2010. 1 Lima: Ed. Imprenta A.G. Carrasco, p. 188, 2017.). During the JFM season, the hydrographic units located in the southern and eastern of the Basin show fluctuation and slope-index negative, which indicates decrease precipitation concerning the average. These Basins are Huancané, Ilave, Suches, Callaccame, Mauri Chico, and the Inter-Basins of 0157, 0173, 0171, 0155, 0175, and Ramis. Ramis shows the lowest slope-index with -0.052, followed by the Huancané with -0.033. In contrast, the hydrographic units located in the north and west show fluctuation and slope-index positives, which indicates an increase in the precipitation concerning the average. The trend analysis from the Mann-Kendall test on the rainfall with a 95% of confidence level not show statistical significance in any period of the time series analyzed.

3.2. Drought hazard

Figure 6 shows the districts with the highest fluctuation of the Standardized Precipitation Index (SPI) for a time scale of 3-months in the period 1981 to 2016. In the Fig. 6.a the 3-months SPI for rainy season (JFM) is analyzed and not the winter period because a significant dry period during the winter season may have few impacts (Wilhite, 2000WILHITE, D.A. Drought as a natural hazard: Concepts and definitions. Published in Drought: A Global Assessment, v. I, p. 3-18, 2000.). During JFM is not detected significant changes in the SPI fluctuation, so not show statistical significance in any period of the time series analyzed. In 1983, 1987, 1990 and 1992, occurred extreme (SPI ≤ -2.00), severe (-1.50 ≤ SPI ≤ -1.99) and moderate (-1.0 ≤ SPI ≤ -1.49) drought events with categories and intensities different. The longest extreme drought occurring in 1983 with a period of 60 days, while the severe droughts in 1990 and 1992 lasted 37 and 40 days. The SPI reached values of -3.76 in 1983 in Vilavila district; -2.82 in 1990 in Antauta district; -2.69 in 1987 in Orurillo district and -2.62 in 1992 in San Antonio district. Severe and moderate droughts were detected in 1982, 1989, 1998, 2005, 2006, 2009 and 2016 in districts of the high-Andean region of Puno, and only moderate drought were observed in 1988, 1991, 1993, 1995, 1996, 2000, 2003, 2008, 2010, 2011, 2013, 2014 and 2015. According to ANA (2013), the drought event in 2011 affected various departments, such as Arequipa, Cajamarca, Lambayeque, Piura, La Libertad, Lima, Moquegua, Tacna, Amazonas, Huánuco, San Martín, Junín and Puno. The drought recorded in 1982-83 resulted from a rainfall deficit during El Niño conditions (analyzed in next section) when upper-level west wind anomalies inhibited the influx of moisture from the east (Sulca et al., 2018SULCA, J.; TAKAHASHI, K.; ESPINOZA, J.C.; VUILLE, M.; LAVADO-CASIMIRO, W. Impacts of different ENSO flavors and tropical Pacific convection variability (ITCZ, SPCZ) on austral summer rainfall in South America, with a focus on Peru. International Journal of Climatology, v. 38, n. 1, p. 420-435, 2018.), causing losses in the Puno region in agriculture that were much larger than in livestock (Claverías, 2016CLAVERíAS, R. El Desarrollo Sostenible en la Cultura Andina. 1. ed. Puno: Universidad Nacional del Altiplano, p. 269, 2016.). Likewise, in 1992 drought coincidentally occurred with El Niño year and was the most severe and affected 16 departments in Peru: Amazonas, Cajamarca, La Libertad, Ancash, Junín, Huánuco, Huancavelica, Pasco, Lima, Cusco, Apurímac, Ayacucho, Arequipa, Tacna, Moquegua and Puno (Endara et al., 2019ENDARA, S.; ACUñA, J.; VEGA, F.; FEBRE, C.; CORREA, K; et al. Caracterización Espacio Temporal de la Sequía en los Departamentos Altoandinos del Perú (1981-2018). Lima: Servicio Nacional de Meteorología e Hidrología del Perú, p. 31, 2019.). On the other hand, it should be noted that in 1983 and 1992 droughts followed a volcanic eruption in the Tropic, El Chichón in April 1982 in Mexico, and Mt. Pinatubo in June 1991 in the Philippines (Imfeld et al., 2019IMFELD, N.; BARRETO, C.; COOREA, K.; JACQUES-COPER, M.; SEDLMEIER, K.; et al. Summertime precipitation deficits in the southern Peruvian highlands since 1964. International Journal of Climatology, v. 39, p. 4497-4513, 2019.).

Figure 6.b shows the precipitation deficit analyzed with the 3-months SPI in every month of the year, finding 59 districts out of 90 (59/90) present negative fluctuation and 31/90 districts present positive fluctuation for the period 1981 to 2016. In the analysis of the SPI, it has been registered 19,317 drought events during the period from 1981 to 2016, of which: 68% (13,074) is slight drought; 18% (3,535) is moderate drought; 9% (1,788) is severe drought, and 5% (920) is extreme drought.

According to Caviedes (1973)CAVIEDES, C. Secas and El Niño: Two simultaneous climatical hazards in South America. In: Proceedings of the Association of American Geographers, v. 5, p. 6, 1973., ENSO and drought are indeed the consequence of the same large-scale variation in the tropical and equatorial circulation, and that the closeness of the years of their occurrence is not just a coincidence. The climate of South America (SA) has long held an intimate connection with El Niño (Cai et al., 2020CAI, W.; MCPHADEN, M.; GRIMM, A.; RODRIGUES, R.; TASCHETTO, A.; et al. Climate impacts of the El Niño-Southern oscillation on South America. Nature Reviews Earth & Environment, v. 1, p. 215-231, 2020. doi
doi... ). The main climate anomalies that characterize the El Niño in South America are related to a modification of the Walker circulation and to a southern shift of the Intertropical Convergence Zone, they include manifestation of the significant rainfall deficits in the Altiplano of southeastern Peru and Bolivia, the Brazilian northeastern, northtern of South America and southern of Central America (Ortlieb and Macharé, 1993ORTLIEB, L.; MACHARé, J. Former El Niño events: records from western South America. Global and Planetary Change, v. 7, n. 1-3, p. 181-202, 1993.). In Peru, El Niño phenomenon mainly generates climatic alteration that is manifested in intense rains in the north and severe droughts in the altiplano region of the south (Fig. 7) where is the hydrographic drainages of Lake Titicaca that present a significant rainfall deficit during the El Niño phenomenon (Lagos et al., 2008LAGOS, P.; SILVA, Y.; NICKL, E.; MOSQUERA, K. El Niño related precipitation variability in Perú. Advances in Geosciences, v. 14, p. 231-237, 2008. doi
doi... ; Lavado and Espinoza 2014LAVADO, W.; ESPINOZA, J.C. Impactos de El Niño y La Niña en las lluvias del Perú (1965-2007).Revista Brasileira de Meteorologia, v. 29, n. 2, p 171-182, 2014.).

Figure 7 shows the correlation between PISCO-precipitation for the JFM season and the C, E, and TNA index. Precipitation extremes in the southern Andean region are highly impacted by SST anomalous conditions in the central and eastern Pacific, showing negative correlations (precipitation deficit) (Figs. 7.b and 7.c). On the other hand, impacts of anomalous conditions in the tropical Atlantic, during the rainy season, is barely detected (Fig. 7.a). These results are in agreement with findings observed in Fig. 6, where SPI peaks (positive and negative) are strongly associated with El Niño and La Niña events recorded in recent decades. Table 5 shows ENSO influence on hydrographic units of the Titicaca Lake Basin through the E (eastern El Niño) and C (central El Niño) index based on Fig. 7. Table 1 shows the districts that are within the hydrographic units. The Basins of Azángaro, Callaccame, Coata, Ilave, Ilpa, Mauri, Mauri Chico, Pucará and Inter-Basins of 0155, 0157, 0173, 0175, present a very strong correlation with the C index which indicates correlation with cold, moderately-warm, and neutral events in the central Pacific, and the Basin of Huancané presented more correlation with the E index, which indicates warm events in the equatorial Pacific.

Figure 8 shows the probability of drought occurrence (PO) in drought categories: moderate, severe, and extreme, in the districts of the Titicaca Lake Basin. The PO levels are low, moderate, high, and very high, and according to each category of drought, the following has been found:

• Moderate drought: In this category are 23 districts out of 90 (23/90) that present very-high PO; 26/90 have high PO; 23/90 maintain a moderate PO and 18/90 have a low level of PO. The districts with high to very-high PO are located in the northeast, central, and southeast of the Titicaca Lake Basin, covering 54% of the Basin.

• Severe drought: In this case, show 23 districts out of 90 (23/90) that present very-high PO; 25/90 have high PO; 22/90 maintain a moderate PO and 20/90 have a low level of PO. The districts with high to very-high PO are located in the north, west, and south-east, and represent 49% of the Basin.

• Extreme drought: In this analysis, it has been found that 27 districts out of 90 (23/90) present very-high PO; 26/90 have high PO; 17/90 maintain a moderate PO and 20/90 have a low level of PO. The districts with high to very-high PO are located in the northwest, northeast, and center-south, and covers 68% of the Basin.

Figure 9 shows the drought hazard of the Titicaca Lake Basin, at the district scale for a time scale of 3-months SPI (in every month), obtained using Table 3. The results suggest that 37 districts out of 90 (37/90) show very-high drought hazard, 9/90 districts present high hazard, and 24/90 show moderate hazard. The districts with a high degree of exposure to drought events (high to very-high hazard) predominate in the northeast and center-south of the Titicaca Lake Basin, covering 51% of the area. In addition, the districts located in the part of the center-north and south of the Basin show low to moderate risk except for Capazo district. The drought hazard is caused by the precipitation deficit that can be associated with the occurrence of special climatic conditions.

3.3. Vulnerability to drought

Figure 10 shows the vulnerability to drought of socioeconomic and physical indicators at the district scale of the Titicaca Lake Basin. The results are discussed below:

• Figure 10.a shows the population density at the district scale. The districts located in the center-east and southeast (except Pisacoma district) present high to very-high population density and cover 29% of the Basin. In this area are the districts of Juliaca and Puno that show a high population density.

• Figure 10.b shows the Economically Active Population (PEA) in the agriculture, livestock, hunting, and forestry sectors. The districts with high to very-high vulnerability are mainly located in the north, center-east, and southeast, covering 60% of the Titicaca Lake Basin. In this area are the larger territorial surface of districts such as Santa Rosa, Nuñoa and Acora.

• Figure 10.c shows the Human Development Index (IDH). The districts located in the northeast and south of the Basin have high to very-high vulnerability and occupy 43% of the area. In this zone is the Carabaya province, which according to the National Institute of Statistics and Informatics (INEI) presents a higher percentage of poverty, as the districts of Crucero and Usicayos.

• Figure 10.d shows the relationship between women and men. The districts located in the center-north have a larger population of women than men, occupying 38% of the Titicaca Lake Basin. According to the Intergovernmental Panel on Climate Change (IPCC), the climate crisis affects women more than men. The provinces with the largest number of districts within this zone are Azángaro, Carabaya, Moho, and San Román.

• Figure 10.e shows livestock production. Most of the districts of the Basin of Lake Titicaca present high to very-high livestock production, covering 80% of the Basin. According to Ardiles (2015)ARDILES, J. PUNO: Planificación, Estrategia de Crecimiento Económico y Desarrollo Regional. 1. ed. Puno: Universidad Nacional del Altiplano, p. 374, 2015., the Puno department is the most important in the country in terms of livestock production and the main domestic supplier of meat. The districts with the highest risk are: Acora, Ilave, Nuñoa, Santa Lucia, Santa Rosa, Juli, Pichacani, Lampa and Azángaro. In addition, the districts with the least number of livestock units are located in the eyebrow of the jungle, and in the southern such as Cuturapi, Unicachi and Tinicachi.

• Figure 10.f shows the rainfed agricultural area, finding that the districts located in the center-north and southeast (except the districts of Capazo, Pisacoma, Huacullani) of the Titicaca Lake Basin, have high to very-high vulnerability to drought and occupy 45% of the area. The main provinces are Puno, Azángaro, Melgar, Chucuito, Huancané, Lampa, El Collao, and San Román.

Figure 11 shows vulnerability to drought at the district scale in the Titicaca Lake Basin. The map is divided into districts with low, moderate, high, and very-high vulnerability obtained 33 districts out of 90 (33/90) with very-high vulnerability, 13/90 districts with high vulnerability, and 22/90 with moderate vulnerability. The districts with high and very-high vulnerability are located in the center-north and southeast of the Basin, occupying 47% of the surface. The Puno region is considered the main producer of livestock in the departments of the high-Andean zone of Peru, however, still maintains levels of monetary poverty higher than the national average (Paredes and Escobar, 2018PAREDES, R.; ESCOBAR, F. El rol de la ganadería y la pobreza en el área rural de Puno. Journal of High Andean Research, v. 20, n. 1, p. 39-60, 2018.). The districts within the Titicaca Lake Basin that present high levels of poverty at the national level, higher livestock production, and depend on the agricultural sector in special rainfed agricultural are Macari, Zepita, Vilque chico, Asillo, Orurillo and Saman. In addition, in the zone of greatest vulnerability are the provinces with the highest population density as Puno, Huancané, Azángaro, San Román, and important districts such as Puno and the city of Juliaca which is considered a gigantic permanent market and is the economic capital of the department of Puno (Apaza, 2016APAZA, H. Actividades Económicas en Juliaca. Revista de Investigación “K´uskiykuy”, v. 1, n. 1, p. 185-194, 2016.). Likewise, the districts of Ayaviri, Muñani, Juli and Azángaro, have a larger surface extension and present more women than men. The districts with moderate to low vulnerability are located in the west, southwest, and northeast. There are areas with a low vulnerability that border areas with high or very high vulnerability, due to the socio-economic and physical indicators are independent in each district.

3.4. Drought risk

Figure 12 shows the drought risk for a time scale of 3-months in the Titicaca Lake Basin at the district scale. The results show that 25 districts out of 90 (25/90) present very-high risk, 20/90 districts with high risk, and 24/90 districts with moderate risk. The districts with high and very high drought risk are localized in the northwestern, center-east, and center-south of the Titicaca Lake Basin, occupying 53% of the Basin. The districts with low to moderate drought risk are located in the zones of southern and center-western of the Basin. The provinces main with a larger number of districts in the zones with high and very-high drought risk are: Azángaro province that present one of the higher population densities and has a larger development in livestock (Romero, 2013ROMERO, E. Monografía del Departamento de Puno. 3. ed. Puno: Universidad Nacional del Altiplano, p. 601, 2013.), and inside are the districts: Achaya, Chupa, Saman, Arapa, Asillo, and San Anton. Another province is Melgar, where are the districts: Orurillo, Macari, Ayaviri, Nuñoa, and Santa Rosa. The Nuñoa district, present the second-largest district area within the Titicaca Lake Basin with high livestock production and agricultural area under rainfed. Additionally, the Carabaya province presents a higher percentage of poverty and contains the districts of Usicayos and Crucero. The Usicayos district is the second poorest district at the national level with a 96.9% incidence of poverty (INEI, 2013INEI. IV Censo Nacional Agropecuario 2012: PUNO Perfil Agropecuario. Punto: Grafía S.A.C., p. 201, 2013.). The provinces that stand out the most for their economy are Puno and San Román. The San Román province has the districts: Cabana, Caracoto and Juliaca. The Puno province contain the districts: Acora, Pichacani, Puno, Mañazo, Plateria, Paucarcolla, Atuncolla, Huata, Chucuito, Capachica and Coata. The Acora district present the third-largest district area within the Titicaca Lake Basin with high livestock production and agricultural area under rainfed. The cities of Juliaca and Puno have a high population density that is 4 and 3 times higher than the departmental average, and concentrate most of the industry, commerce, as well as infrastructure, and state services of the Puno department (Ardiles, 2015). The city of Juliaca is considered as the economic capital of the department of Puno; there local, regional, national and cosmopolitan entrepreneurs compete (Apaza, 2016APAZA, H. Actividades Económicas en Juliaca. Revista de Investigación “K´uskiykuy”, v. 1, n. 1, p. 185-194, 2016.).

4. Conclusions

The drought hazard was estimated as the precipitation deficit monthly rainfall obtained from product PISCO-Precipitation v1.2 (Peruvian Interpolation of the SENAMHIs Climatological and Hydrological Stations) for the period 1981 to 2016 (36 years), and then was quantified with the Standardized Precipitation Index (SPI) for a time scale of 3-months. The rainy season of the Titicaca Lake Basin occurs in January, February, and March (JFM), due to that more than 50% of precipitation accumulates in these months. The rainy months in the Basin are very important because the main economic activities depend on the rains for their development. The hydrographic units of the Basin, located in the southern and eastern show negative fluctuation of precipitation, which indicates a decrease in the precipitation concerning the average, however, not show statistical significance. According to the SPI, in 1983, 1987, 1990, and 1992 in the rainy season occurred extreme, severe, and moderate drought events, with different categories and intensities. Additionally, in 1982, 1989, 1998, 2005, 2006, 2009, and 2016, were detected in some districts of the Basin severe and moderate droughts. The areas with larger predominant of drought hazard are localized in the northeast and center-south of the Titicaca Lake Basin. Our results demonstrated that precipitation deficit occurs mainly associated with the occurrence of anomalous warm SST conditions in the central Pacific rather than the eastern Pacific (Figs. 3, 7 and Table 5).

The Titicaca Lake Basin is located within the high-Andean region of the department of Puno, which is one of the departments with more poverty at the national level. The drought vulnerability varies in the districts due to the differences in socioeconomic and physical indicators, for this reason, the Basin present districts with a low vulnerability that border districts with high or very high vulnerability. The zones with high and very-high vulnerability are located in the center-north and southeast of the Basin. These zones include provinces with the highest population density, districts with the highest levels of poverty, and those economically dependent on the livestock and agricultural sectors, in special rainfed agricultural.

In the Titicaca Lake Basin was identified 45/90 districts with high to very-high risk of drought, that was obtained based on hazard and vulnerability. These districts have a large deficit of precipitation concerning the average and main socioeconomic and physical indicators vulnerable to drought hazard. The districts occupy 53% area of the Basin that are located in northwestern, center-east, and center-south. Within this area are important provinces such as Puno, San Roman, Azángaro, Melgar, and Carabaya, as well as the districts of Puno and Juliaca, the latter considered the economic capital of the department of Puno.

The Titicaca Lake Basin is the poorest region of the country with an economy that depends on precarious rainfed agriculture and an extensive and mixed livestock rearing system, so the impact of droughts generates large economic losses and increases poverty in the region, as has been recorded in the past. Therefore, based on the present study where the districts with the highest risk of droughts are identified, is important that decision-makers and the population can develop preventive measures to mitigate and reduce the impacts of the drought phenomenon in the high-Andean region of Puno.

References

Internet Resources

Publication Dates

History