Beyond Global Trends: Two Decades of Climate Data in the World’s Highest Equatorial City

Abstract

While humanity stands at a critical point—one future leading toward sustainability, equity, and resilience, the other toward escalating conflicts, ecological collapse, and irreversible loss—climate change emerges as one of the most urgent challenges of the 21st century. The Global South, specifically the northwestern South American region, lacks model confidence and reports on current climatic conditions due to gaps in historical data. This study, therefore, presents temperature and precipitation trends in the highest city on the equator, Quito, Ecuador, from 2004–2024. Six different districts were analyzed for maximum, average, and minimum temperatures, as well as cumulative precipitation, in terms of monthly and annual statistics, using Seasonal-Trend Decomposition. Over the past two decades, this Andean city has warmed by an average of +0.95 °C, with minimum temperatures rising at rates twice the global urban average of extreme urban heat islands (+2.47 °C), while precipitation has nearly doubled in rapidly developing parts of the city. These profound changes, shaped by urban expansion, El Niño–Southern Oscillation variability, and climate change, demand urgent adaptation in water management, urban planning, and climate resilience strategies, as well as comparative studies with rural Ecuador to differentiate local vs. regional climate signatures.

1. Introduction

Humanity currently stands at a crossroads between two future futures: one of sustainability, poverty eradication, and technological advancement; or, conversely, one of increasing inequality, armed conflicts, ecological crises, and insufficient progress in addressing the impacts of climate change [1]. The unprecedented surge in global population—an increase of approximately two billion people over 23 years (1999–2022)—combined with the relentless release of anthropogenic greenhouse gases, is driving continuously intensifying global warming, regarded as the most critical health threat facing humanity [2,3,4,5]. While the Paris Agreement aims to limit the global average temperature increase to well below 2 °C (preferably not much over 1.5 °C) compared to pre-industrial levels, by 2024, global surface temperatures had already reached 1.47 °C above pre-industrial values, marking the hottest year on record and signaling how dangerously close we are to breaching the 1.5 °C threshold [6,7,8,9]. Consequently, rising temperatures are triggering a wide range of environmental, climatic, and social impacts, including biodiversity loss, human casualties, damage to infrastructure, threats to food security, population displacement, and reduced electricity production, among others [10].

Although climate change poses a universal challenge, Latin America—and particularly northwestern South America (NWS), which includes Ecuador—exhibits distinct vulnerabilities. With over 80% of its population residing in urban areas, this region is considered one of the most climate-sensitive in the world [11,12]. Complex socio-economic factors such as poverty, inequality, and political and economic instability further heighten the region’s climate vulnerability [13,14]. Despite contributing minimally to global greenhouse gas emissions, countries in the NWS region have experienced a 30–50% retreat of tropical Andean glaciers since the 1980s, leading to reduced water availability, heightened risks of floods and landslides, disrupted hydropower generation, and pronounced shifts in temperature and precipitation regimes [12]. These changes directly affect urban centers such as Quito by increasing flood susceptibility, threatening water supplies, impacting agriculture and hydropower, and accelerating biodiversity loss [13,15,16]. The regional climate stressors outlined above underscore the urgent need for localized studies on temperature and rainfall trends to inform targeted adaptation and resilience strategies.

Previous studies emphasize persistent deficiencies in long-term climate data coverage across many developing regions [17,18]. Within the NWS, few studies incorporate adequate station density, temporal duration, or spatial representativeness, resulting in low confidence in future projections for this area. Limited observational networks and non-representative sampling introduce substantial uncertainties into climate forecasts, particularly regarding shifts in precipitation extremes [12]. Temperature trends show rates exceeding global averages across several South American sub-regions, with warming particularly pronounced in the NWS [12]. At the same time, Andean glaciers continue to experience some of the most significant mass losses worldwide [19,20,21,22]. Model projections suggest potential reductions in precipitation of 20–30% in Ecuador and nearby portions of the northern tropical Andes [23,24], which would place additional stress on water resources and agriculture [12]. The high climatic sensitivity of the tropical Andes, combined with substantial microclimatic variability over short distances, suggests that broad generalizations may obscure critical local adaptation needs.

The Ecuadorian capital, Quito, represents a relevant case study due to its rapid demographic and spatial transformations [25]. Urban expansion into surrounding valleys and increased densification have likely intensified urban heat island effects, driven by higher pollutant emissions, deforestation, and the spread of impervious surfaces [26]. These dynamics, combined with the city’s elevation and complex topography, likely influence local climate responses in ways that remain insufficiently documented. Global temperature datasets often adjust urban records to remove the urban heat island effect, meaning that cities such as Quito are not fully represented in their unadjusted form. This can hinder assessments of urban climate vulnerabilities [27]. The scarcity of publications from Central and South America does not reflect reduced exposure or sensitivity to climate change impacts, but is instead largely attributable to structural factors such as limited long-term observations and institutional capacity [28].

Given that global surface temperatures have risen at an unprecedented pace, with rates accelerating markedly over the past 24 years [2,10], this study seeks to address critical data gaps by quantifying temperature and precipitation trends in Quito from 2004 to 2024. It examines their links to global climate drivers (e.g., El Niño–Southern Oscillation, ENSO) and local factors such as urban expansion and land-use change, thereby providing essential high-resolution insights to inform adaptation strategies in high-elevation tropical urban environments.

2. Materials and Methods

2.1. Study Site

Quito, the capital of Ecuador, is often cited as the world’s highest constitutional capital, standing at 2850 m above sea level (ASL) (Figure 1) [29]. Its metropolitan area of 4,183 km² spans about 95 km along its north–south axis and about 40 km east–west. This territory encompasses a substantial elevational range, crossing the equator and intersected by two principal Andean mountain chains: the Eastern Cordillera, which rises to elevations near 4100 m ASL, and the Western Cordillera, reaching up to around 4800 m ASL [30]. This combination of topographical complexity results in a range of tropical climatic zones and natural ecosystems [30].

Quito’s population now exceeds 2.7 million residents, effectively doubling within the past twenty years and reflecting substantial urban growth and transformation of its surrounding landscapes [31]. Despite significant socio-economic and environmental constraints, Quito maintains a generally temperate, spring-like climate, characterized by two primary seasons: an extended wet period from September to May and a drier interval from June to August [32]. The temperature in this region shows little seasonal variation, maintaining an annual mean of approximately 13.4 °C, although it is marked by substantial diurnal fluctuations, with daytime highs reaching around 29.9 °C and nighttime lows dropping to approximately 1.5 °C [33,34]. Finally, the city’s high-altitude location results in elevated solar and UV radiation [35].

For this work data from six meteorological stations representing different districts of Quito were investigated: (1) Carapungo (2660 m ASL, coord. 78°26′50″ W, 0°5′54″ S); (2) Cotocollao (2739 m ASL, coord. 78°29′50″ W, 0°6′28″ S); (3) Belisario (2,835 m ASL, coord. 78°29′24″ W, 0°10′48″ S); (4) Tumbaco (2,355 m ASL, coord. 78°24′12.4164″ W, 0°12′53.334″ S); (5) Camal (2,840 m ASL, coord. 78°30′36″ W, 0°15′00″ S); and (6) Chillos (2453 m ASL, coord. 78°27′36″ W, 0°18′00″ S) (Figure 1 shows the spatial distribution of the stations). Study sites 4 and 6 are located in recently rapidly developing suburban, lower-elevation valley districts.

2.2. Data Acquisition and Analysis

The municipal office, the Secretaría de Ambiente de Quito, has operated the atmospheric monitoring network since 2004. Air temperature was measured using a Thies Clima Hygro-Thermo Transmitter compact sensor (Model 1.1005.54.161). Precipitation was measured using a Thies Clima tipping bucket rain gauge (Model 5.4003.2007) (Adolf Thies GmbH & Co. KG, Göttingen, Germany). Data acquisition and storage were performed with an Agilaire 8872 data logger (Agilaire, LLC, Knoxville, TN, USA). These instruments provided continuous measurements essential for analyzing local climatic variability. The collected data are stored in the municipal administration’s repository and are available for download at https://datosambiente.quito.gob.ec/ (accessed on 20 May 2025).

The analysis of temperature and precipitation trends in Quito, Ecuador, from 2004 to 2024 was conducted using a combination of data extraction, processing, and statistical techniques. The AirEcuador v0.2 [36] package was used for data retrieval, and the OpenAir v2.19.0 [37] was employed for time-series analysis.

The dataset was extracted from the Secretaría de Ambiente de Quito’s repository and filtered to cover the period from 1 January 2004 to 31 December 2024, ensuring a consistent 21-year timeframe. Two sites contain incomplete datasets: (4) El Camal, which spans 1 January 2004 to 31 December 2022; and (6) Tumbaco, which begins in October 2006. Data loss across all stations was less than 5%, which was deemed sufficient for analysis. Missing values and potential infinities were removed to ensure data integrity, with days containing incomplete records (i.e., <75% of daily data or fewer than 18 of 24 h) excluded from monthly and annual statistics. The raw data underwent quality control using the AirEcuador package, which removed anomalous values at the 99.9th percentile, as the Secretaría de Ambiente repository provides unprocessed data. Subsequently, the data were transformed into a long-format structure using the tidyr v1.3.1 package, with precipitation values aggregated monthly using the dplyr v 1.1.4 package. To ensure comparability across sites, analyses focused on overlapping data periods where available, maximizing the use of the 2004–2024 timeframe.

Monthly precipitation sums were calculated for each station, and annual totals were derived using the timeAverage function from the OpenAir package. To explore long-term trends, a Seasonal-Trend Decomposition using LOESS (STL) was applied to the monthly time-series data for each station [38]. The STL decomposition, implemented using the stl function from the forecast package [39], employed a periodic seasonal window and robust fitting to extract the trend component, thereby accommodating nonlinear changes over the 21-year period. The isolated trend component was then evaluated using the Mann–Kendall test to assess statistical significance, while total changes (Δ) over the study period were derived directly from differences in the STL trend.

This approach isolates the nonlinear trend component while preserving the data’s absolute scale (°C or mm), thereby facilitating direct interpretation of long-term shifts in typical temperature and precipitation levels. Because the primary objective was to quantify gradual changes in the baseline climate rather than transient departures from expected seasonal patterns, anomaly analysis was not performed. The STL decomposition effectively accounts for periodic seasonality, allowing robust estimation of multi-decadal trends without requiring explicit conversion to anomalies.

The graphical representation was constructed using ggplot2 [40]. The STL trend lines were overlaid to illustrate the smoothed long-term trajectory. To quantify changes, the percentage delta (Δ) was calculated for each station, representing the relative change in the STL trend between the base period and December 2024. The baseline was defined as the STL-derived trend value in 2004 (or the first available year of data for stations with shorter records), and the final value was taken as the STL trend for December 2024 (or the last available year of data). The Δ was computed as ((Trend_final − Initial_value)/Trend_base) × 100, with the results annotated on the graphs. The total change (Δ) over the period is reported with 95% confidence intervals (CI), computed using the bootstrap method, representing, together with the p-value, the statistical significance of the trend (Mann–Kendall test).

The influence of ENSO was assessed by integrating the Oceanic Niño Index (ONI) from NOAA

(https://origin.cpc.ncep.noaa.gov/products/analysis_monitoring/ensostuff/ONI_v5.php (accessed on 5 May 2025)).

Monthly ONI 3.4 values were compared with citywide averages of monthly temperatures and precipitation accumulation.

Satellite images to observe land-cover changes in the Quito region were retrieved from Google Earth for three stages representing the study period: (i) 12/31/2002; (ii) 12/31/2013; and (iii) 1/1/2021–6/11/2024, by shifting the historical tool on the Google Earth website. Different satellite dataset were used to produce the maps: Landsat/Copernicus (i,ii) and AirbusCNES/AirbusMaxar TechnologiesLandsat/Copernicus (iii) (retrieved from: https://earth.google.com/web/@-0.1927554,-78.3637257,2472.83099683a,68795.46412996d,35y,0.09904646h,0t,0r/data=CgRCAggBOgMKATBCAggASg0I____________ARAA (accessed on 18 July 2025)).

3. Results and Discussion

3.1. Temperature Trends in Quito over 2004–2024

Temperature pattern analysis (with seasonality removed), using average, minimum, and maximum monthly time series, revealed a consistent warming trend at all six sites in Quito, except for maximum temperature at two northern sites: (1) Carapungo and (2) Cotocollao (Figure 2). Statistically significant (p < 0.01; CI: 95%) increases in average monthly temperatures ranged from +0.63 ± 0.26 °C to +1.45 ± 0.31 °C, equivalent to annual rates between +0.30 °C/decade and +0.79 °C/decade, with a city-wide average of +0.52 °C/decade (Figure 2, black lines). Minimum monthly temperatures show even more pronounced significant increases (p < 0.01), ranging from +1.45 ± 0.37 °C to +2.57 ± 0.51 °C, or +0.69 °C/decade to +1.41 °C/decade, averaging +1.02 °C/decade city-wide (Figure 2, blue lines). Meanwhile, changes in maximum monthly temperatures varied from −0.51 ± 0.51 °C to +0.87 ± 0.47 °C, yielding a city-wide mean of +0.12 °C/decade. Excluding non-significant trends in (1) Carapungo, (2) Cotocollao, and (5) Camal, this rose to +0.33 °C/decade (Figure 2, red lines). Notably, although two northern sites (Sites 1 and 2) exhibited negative slopes for maximum temperatures, these were not statistically significant (p ≥ 0.05; CI:95%), indicating no robust directional change in those districts.

Analysis of annual data supports these results, indicating that the mean temperature across Quito has increased by 1.2 °C over the past 21 years, corresponding to an average rate of 0.57 °C per decade (Figure 3). This closely matches the monthly trend of +0.95 °C, confirming consistency across temporal scales.

The observed changes in average temperatures align with globally reported urban warming rates. However, minimum monthly temperatures in Quito have increased at roughly twice the global rate for urban areas. Meanwhile, maximum temperatures are much lower than globally reported values. For comparison, a recent global analysis (2002–2021) reported increases of +0.50 ± 0.20 °C/decade for average temperatures, +0.56 ± 0.21 °C/decade for minimum temperatures, and +0.43 ± 0.16 °C/decade for maximum temperatures [41]. Our results closely resemble the averages reported by the global study on extreme urban heat islands, where temperatures have risen by ~1 °C since 2003 [42]. In our study, Quito’s average temperature increased by 0.95 °C over a comparable period. We note, however, that the (4) Tumbaco site, which exhibited the highest warming trend, had limited early-year data availability, which may have influenced the magnitude of the detected trend (Figure A2, Appendix A).

These elevated rates likely reflect the combined effects of rapid, often unregulated urban expansion, limited investment in climate-adaptive infrastructure, and weaker environmental governance, patterns commonly observed in cities of the developing world [41,43]. Over the past two decades alone, Quito’s population has more than doubled, from 1.3 to 2.7 million, accompanied by substantial growth in the vehicle fleet. This growth coincided with major new road construction toward the eastern valley (notably, Tumbaco), adding nearly 30 km of paved infrastructure between 2011 and 2015. Urban expansion has often replaced rural lands and led to densification of urban and suburban zones, where traditional single-family homes with gardens have been replaced by high-rise residential and office buildings.

Among the study sites, the rapidly developing peripheral district (4) Tumbaco recorded the highest increase in average and minimum monthly temperatures over the last 18 years, +1.45 °C and +2.57 °C, respectively (Figure 2d and Figure A2a, Appendix A). Therefore, the actual long-term warming may be underestimated. Finally, the largest changes in maximum temperatures were recorded in both rapidly developing peripheral lower valley districts, ranging from +0.68 °C to + 0.87 °C, respectively. These findings are consistent with an intensifying urban heat island (UHI) effect [26]. In these two districts, a rapid conversion of natural ecosystems and farmland into residential developments is evident (Figure A1, Appendix A). Here, the probability of transition from vegetated to impervious surfaces is particularly high [44], which has implications for both the water cycle and urban albedo [45]. This UHI effect is likely compounded by reduced vegetation cover and increased heat absorption from impervious surfaces. Conversely, more established central districts such as Belisario, which maintain greater green space, show a more moderate UHI impact (Figure 2c), emphasizing the role of urban greening in mitigating temperature rise. This agrees with previous findings that while green infrastructure provides cooling benefits, it does not fully offset UHI intensification [46,47].

Residential growth has also extended into surrounding valleys and up the slopes of the metropolitan area, frequently taking place outside formal planning frameworks and, in many cases, violating land-use regulations [48]. In contrast, the historic center of Quito has seen limited physical expansion due to land scarcity, high population density, traffic congestion, and aging infrastructure. As a result, urban growth has increasingly shifted into nearby valleys, placing additional pressure on vital resources such as water supplies and agricultural lands. As this study demonstrates, this pattern of expansion is also associated with notable local temperature increases, raising important concerns about the long-term sustainability of these developments [49].

3.2. Precipitation Trends in Quito over 2004–2024

Two southern sites, (5) El Camal (Figure 4e) and (6) Los Chillos (Figure 4f), along with the central site (3) Belisario (Figure 4c), ranked among the wettest districts, with average annual totals of 975 mm, 1,037 mm, and 998 mm, respectively. In contrast, northern sites (1) Carapungo (Figure 4a) and (2) Cotocollao (Figure 4b) recorded lower annual means of 580 mm and 688 mm, reflecting generally drier conditions, while (4) Tumbaco (Figure 4d) in the eastern valley was the driest at 657 mm/year.

Despite these differences in baseline precipitation, the rates of increase vary substantially across stations (Figure A2, Appendix A). Over 2004–2024, (5) El Camal exhibited a moderate increase, Δ = 10.16% of the base trend, while (4) Tumbaco showed the largest relative increase, Δ = 105.32% (Figure 4 and Figure A2, Appendix A). (2) Cotocollao and (3) Belisario recorded notable increases of Δ = 81.20% and Δ = 85.80%, respectively, with (1) Carapungo rising by Δ = 34.55% and (6) Los Chillos by Δ = 60.08%, suggesting a recent shift from earlier trends driven by regional moisture dynamics. Such changes may be driven by urban-induced modifications to local moisture dynamics and convection [50]. The relatively small change at (5) El Camal, despite its high annual totals, could be due to its location combined with data limitations, whereas the larger increases at (6) Los Chillos and (3) Belisario reinforce the pattern of a wetter south-central gradient.

As revealed by STL trend decomposition, annual precipitation trends across Quito show clear spatial differences along with an overall increase in cumulative precipitation over the past 21 years. Some areas of Quito have experienced extraordinary increases in annual precipitation, nearly doubling over the past two decades. Such sharp rises are exceptional and may signal shifts in regional climate dynamics, urban-enhanced convection linked to higher temperatures [51,52], or amplified effects of ENSO variability. These trends have critical implications for flood risk, water infrastructure, and land-use planning. They underscore the urgency of integrating climate adaptation strategies into urban development, considering not only long-term precipitation changes but also event frequency and intensity through methods tailored to these highland zones [53].

Regionally, in the Northern Tropical Andes, recent studies have confirmed increases in the intensity and frequency of precipitation events over the past decade, strongly modulated by ENSO and exacerbated by climate change, which also introduces greater unpredictability [54]. Additional research has documented rising relative humidity trends in northern Ecuador and southern Colombia [55]. These results also support broader findings reported by the IPCC, which indicate with medium confidence that globally averaged precipitation over land has increased since 1950, with an accelerated rate of change since the 1980s [2].

3.3. El Niño Southern Oscillation Influence on Quito’s Temperature and Precipitation

Figure 3 illustrates the influence of ENSO events on local temperatures. Almost every El Niño year (marked by red symbols) corresponded to a new local temperature maximum, exceeding the long-term warming trend line. Conversely, La Niña years (blue symbols) generally showed cooler conditions, falling below the trend. While the inter-Andean region of the Ecuadorian Andes is characterized by pronounced diurnal temperature ranges, these variations, when averaged over days and months, result in relatively modest seasonal fluctuations. This suggests that more substantial interannual temperature variability is linked to global or regional climate phenomena, overlaid on longer-term warming trends associated with climate change. Temperature changes across the Andes have long been connected to ENSO dynamics, particularly through teleconnections with sea surface temperature anomalies in the central Pacific’s Niño 3.4 region, which are known to also influence drought conditions [56].

This pattern aligned with the well-established influence of ENSO on precipitation in the Andes, where El Niño typically reduces rainfall and La Niña enhances moisture availability [56] (Figure 5). In Quito, this variability has direct implications for water management: El Niño years may intensify wildfire risk and water scarcity in drier districts such as (4) Tumbaco, while La Niña phases can increase flood hazards, including landslides, debris flows, and urban flooding in wetter areas such as (5) El Camal and (6) Los Chillos. These findings underscore the need for adaptive strategies, favoring nature-based or hybrid solutions that address both long-term trends and ENSO-driven variability to build resilient urban planning and water systems.

In the Ecuadorian Andes, El Niño patterns were observed on glaciers on nearby volcanoes [57], as well as in the shifting elevations of glacier snow lines during different ENSO phases [58]. These relationships are driven by the propagation of El Niño conditions through the mid-troposphere, generating upper-level high-pressure systems that suppress rainfall and promote drought in the Andes [56]. Moreover, the recent 2023–2024 El Niño event highlighted these connections even more starkly, exhibiting pronounced anomalies in both precipitation and temperature patterns in the region [59,60] (Figure 3 and Figure 5). This may signal a trend toward more frequent and severe climate extremes in areas teleconnected to ENSO activity, including Ecuador and the broader South American tropics. Moreover, climate change may be amplifying the impacts of such events, prolonging thermal anomalies well beyond the typical lifespan of an El Niño episode [61]. In Ecuador, these compounded effects were evident in cities such as Quito, which faced heightened wildfire risks, hydroelectric power shortages, and increased strain on drinking water supplies [13].

3.4. Synthesis

In this work, we provide a uniquely detailed, multi-site analysis of temperature and precipitation trends within a high-elevation tropical capital on the equator, revealing patterns that exceed global urban averages in rapidly developing parts of the city. While previous studies in the Tropical Andes have focused on regional or altitudinal gradients or examined single stations over shorter timeframes, our work captures fine-scale intra-urban variability across various districts in Quito over two decades. It documents exceptional rates of minimum temperature increases—up to twice the global urban average—and localized precipitation rises exceeding 100%, underscoring the combined influence of rapid urban expansion, Andean topography, and teleconnections with ENSO events. By integrating high-resolution temporal and spatial data, this study not only fills critical observational gaps for northwestern South America but also highlights how the convergence of global climate change and unplanned urbanization can intensify climate impacts in tropical highland cities. These insights provide an essential foundation for designing localized adaptation and resilience strategies tailored to the unique challenges faced by rapidly growing equatorial metropolises. Results will also serve to inform Quito’s Climate Action Plan 2030 and urban greening initiatives aimed at mitigating UHI and flood risk.

4. Conclusions

This study reveals that tropical high-elevation cities such as Quito are experiencing profound and uneven climatic transformations, driven by the combined influence of global climate change, complex Andean topography, and rapid urbanization. Over 2004–2024, average temperatures in Quito rose by +0.95 °C (0.71 °C/decade), with minimum temperatures increasing by up to 2.57 °C (1.41 °C/decade) in some districts—roughly double global urban warming trends. These findings emphasize a pronounced Urban Heat Island effect, especially in rapidly expanding suburban areas where natural landscapes have been replaced by impervious surfaces.

Precipitation has also intensified, with several districts experiencing increases exceeding 80–100% over two decades, effectively doubling rainfall in some areas. Such sharp rises signal substantial changes in local hydroclimate and may reflect more frequent or intense extremes, although localized factors and data gaps remain.

The El Niño–Southern Oscillation continues to modulate interannual variability, with El Niño years bringing hotter, drier conditions and La Niña phases enhancing wetness and flood risk. The 2023/24 El Niño exemplifies how climate change may be amplifying these impacts, exacerbating droughts, water shortages, and wildfire risks.

The findings of this study stress the urgent need for high-resolution, site-specific climate assessments in tropical highland cities. Understanding how urbanization, topography, and global climate forces interact is essential for developing effective adaptation and resilience in similarly vulnerable regions.