Climate-driven spatiotemporal dynamics of Aedes infestation and dengue transmission in Porto Alegre, Southern Brazil.

This study analyzes spatiotemporal data from Porto Alegre (2018–2025) to demonstrate how climatic variables, particularly rainfall frequency and lagged temperature, drive *Aedes* mosquito infestation and dengue transmission, ultimately validating predictive models that integrate entomological and environmental surveillance to improve public health interventions.

The Gist

🌧️ The Big Picture: When the Weather Turns the City into a Mosquito Hotel

Imagine Porto Alegre, a city in Southern Brazil, as a giant hotel. For a long time, this hotel was mostly empty of a specific guest: the mosquito (specifically Aedes aegypti and Aedes albopictus). Because the weather there used to be a bit cooler and less predictable, these mosquitoes didn't feel welcome, and the virus they carry (Dengue) stayed away.

But recently, the "weather manager" of the city has been changing the rules. With climate change bringing hotter temperatures and weird rain patterns, the hotel is suddenly full of mosquitoes. This study is like a detective report trying to figure out exactly how the weather is filling the hotel and when the guests (people) are likely to get sick.

🔍 The Detective Work: What They Did

The researchers acted like weather detectives and mosquito hunters. They looked at data from 2018 to 2025 (a very recent and crucial time) to answer three main questions:

1. Where are the mosquitoes hiding?
2. What weather conditions make them multiply?
3. When will the mosquitoes bite enough to cause a Dengue outbreak?

They used high-tech traps (MosquiTRAPs) to catch mosquitoes, checked them for the virus, and compared this data with daily weather reports and hospital records of Dengue cases.

🌡️ The Big Discoveries (The "Aha!" Moments)

1. The "Rainy Day" vs. "Rainy Bucket" Surprise

You might think that the total amount of rain (like a giant bucket pouring water) is what makes mosquitoes happy.

• The Reality: The study found that how often it rains matters much more than how much it rains.
• The Analogy: Imagine a garden. If you dump a whole bucket of water on it once a month, the plants might drown or the soil might wash away. But if you give it a gentle sprinkle every day for a week, the plants (and mosquito larvae) thrive.
• The Finding: For the main mosquito (Aedes aegypti), having rain on 4 out of 5 days was the sweet spot for population explosions. It's not about the flood; it's about the consistent dampness that keeps their breeding pools from drying out.

2. The Temperature Time Machine

Mosquitoes are cold-blooded; they move and grow faster when it's warm.

• The Finding: The researchers found a "time delay." If the temperature spikes today, the mosquito population doesn't explode tomorrow. It takes about 2 to 4 weeks for those warm days to turn into a swarm of biting mosquitoes.
• The Analogy: Think of it like baking a cake. You turn on the oven (the heat) today, but you don't get the cake (the mosquitoes) until 30 minutes later. If you see a heatwave, you know the "mosquito cake" will be ready in a month.

3. The Two Types of Mosquitoes

There are two main suspects:

• The City Dweller (Aedes aegypti): Loves the city center, crowded neighborhoods, and artificial containers (like old tires or buckets). This one is the main driver of Dengue.
• The Nature Lover (Aedes albopictus): Prefers parks, gardens, and greener areas. While they are present, they are less efficient at spreading Dengue in this specific city compared to their city-dwelling cousins.

4. The "Hot Zones" Map

The study created a map showing that mosquitoes aren't spread evenly. They form clusters (like groups of friends hanging out in specific neighborhoods).

• Some areas (like the North and East of the city) became "Super Hot Zones" where mosquito traps caught huge numbers.
• These clusters stayed in the same places year after year, acting like permanent mosquito neighborhoods.

📉 The Prediction Machine: The Crystal Ball

The researchers built a computer model (using a method called LASSO) to predict the future.

• How it works: They fed the model past weather data and mosquito counts.
• The Result: The model is surprisingly good at predicting low to moderate mosquito levels. It's like a weather forecast that can reliably tell you, "Hey, in 3 weeks, expect a lot of mosquitoes in this neighborhood."
• The Limitation: It's harder to predict the extreme peaks (the massive outbreaks). Just like a weatherman can predict rain but might miss a once-in-a-century hurricane, the model smooths out the extreme spikes.

🚨 Why This Matters for You

This isn't just academic; it's a public health warning system.

1. Early Warning: Because there is a 4-week delay between hot weather and a mosquito swarm, health officials can use this data to warn people before the outbreak happens.
2. Targeted Action: Instead of spraying the whole city (which is expensive and wasteful), they can send teams to the specific "Hot Zones" identified by the map.
3. Climate Change Reality: The study confirms that Southern Brazil, which used to be safe from Dengue, is now becoming a prime target because the climate is changing. The "season" for mosquitoes is getting longer.

🏁 The Bottom Line

The paper tells us that weather is the conductor, and mosquitoes are the orchestra. When the temperature rises and rain becomes frequent (even if not heavy), the orchestra starts playing louder.

By listening to the weather forecast and watching the mosquito traps, we can predict when the music is about to get too loud (an outbreak) and take action to quiet it down before people get sick. It's about using data to stay one step ahead of the mosquitoes.

Technical Summary

1. Problem Statement

Dengue fever, historically limited in Southern Brazil due to cooler climates, is experiencing a significant expansion driven by climate change, urbanization, and globalization. Porto Alegre, the capital of Rio Grande do Sul, has recently faced extreme weather events (temperature anomalies, flooding, heavy rains) and a surge in dengue cases. Despite this, there is a critical knowledge gap regarding how specific climatic variability influences Aedes vector dynamics and dengue transmission at the local scale in temperate regions. Understanding these drivers is essential for improving early warning systems and public health interventions in areas where dengue was previously considered low-risk.

2. Methodology

The study conducted an integrated spatiotemporal analysis of entomological, epidemiological, and climatic data in Porto Alegre from 2018 to 2025.

• Data Sources:
  • Entomological: Weekly surveillance data from MosquiTRAPs (approx. 1,447 traps) capturing female Aedes aegypti and Aedes albopictus. Metrics included the Mean Female Aedes Index (MFAI) and the DENV-positive trap index (via RT-PCR).
  • Epidemiological: Autochthonous dengue case data from the Notifiable Diseases Information System (SINAN) and InfoDengue platform.
  • Climatic: Daily data from the Copernicus Data Store (ECMWF), including temperature, dew point (converted to relative humidity), and precipitation.
  • Land Cover: MapBiomas data (grouped into anthropic vs. vegetation cover), though these were found to be static during the study period.
• Statistical & Modeling Approaches:
  • Spatial Analysis: Moran's Global Index ($I$) and Local Indicators of Spatial Association (LISA) to identify clustering and spatial autocorrelation of vectors and cases.
  • Correlation Analysis: Kendall's rank correlation ($\tau$) to assess relationships between climate variables and vector abundance, utilizing lagged variables (0–4 weeks).
  • Polynomial Regression: Used to model non-linear relationships between rainfall frequency (number of rainy days) and mosquito abundance across 5, 10, and 30-day windows.
  • Predictive Modeling: LASSO (Least Absolute Shrinkage and Selection Operator) regression was employed to build predictive models for mosquito infestation and dengue incidence. This method handled multicollinearity and selected the most relevant lagged climatic and entomological predictors.
  • Indices: A "Vector-Fitness Infestation Index" (VFII) was created by multiplying MFAI by a climate-driven mosquito-viral suitability Index ($P$).

3. Key Contributions

• Novel Integration: First study to integrate high-resolution entomological surveillance with climatic data to model dengue dynamics in Southern Brazil's temperate zone over a 7-year period.
• Rainfall Frequency vs. Accumulation: Demonstrated that the frequency of rainy days (number of days with >1mm rain) is a stronger predictor of Ae. aegypti abundance than total accumulated rainfall, highlighting the importance of breeding site stability over volume.
• Lagged Predictors: Quantified the temporal lag (up to 4 weeks) between climatic conditions, vector infestation, and subsequent dengue incidence, providing a critical window for early warning.
• Species Differentiation: Provided distinct ecological profiles for Ae. aegypti (dominant, urban, spring-summer peak) and Ae. albopictus (lower abundance, persistent in specific districts, autumn-winter peaks).

4. Key Results

• Spatiotemporal Dynamics:
  • Vector Distribution: Ae. aegypti showed strong spatial clustering that expanded from the city center to the North, East, and Eixo Baltazar regions by 2025. Ae. albopictus remained restricted primarily to the South and Central-South regions.
  • Seasonality: Ae. aegypti peaked in Spring-Summer (March), while Ae. albopictus showed higher captures in Autumn-Winter (March/April), suggesting adaptation to cooler temperatures.
  • Dengue Trends: Dengue incidence increased steadily from 2022 to 2025, with 2025 recording the highest levels. Hotspots shifted over time, with the Eixo Baltazar region reporting the highest burden in 2025.
• Climatic Drivers:
  • Temperature: Showed a strong positive correlation with mosquito abundance, with effects persisting up to 4 weeks.
  • Rainfall: Total accumulated precipitation had minimal influence. However, rainfall frequency (number of rainy days) had a significant non-linear positive effect on Ae. aegypti (peaking at ~4 rainy days in a 5-day window).
  • Humidity: Exerted a weak negative correlation with infestation.
• Predictive Modeling:
  • LASSO Performance: Models demonstrated good agreement between observed and predicted values, particularly for low-to-moderate infestation levels.
  • Key Predictors: Lagged temperature (3–4 weeks) was the strongest climatic predictor for both species. For dengue incidence, the MFAI of both species and the VFII of Ae. albopictus were the most significant predictors.
  • Lag Effect: The correlation between mosquito infestation (MFAI) and dengue cases strengthened progressively up to a 4-week lag, confirming the lead time required for transmission cycles to manifest in case numbers.
• Vector Fitness: The Vector-Fitness Infestation Index (VFII) for Ae. aegypti showed a significant increasing temporal trend ($\beta = 0.0038, p=0.001$), indicating rising transmission potential due to climatic suitability.

5. Significance

• Public Health Intervention: The study validates that integrating entomological surveillance with climatic data can serve as an effective early-warning system. The identified 4-week lag allows health authorities to anticipate outbreaks and deploy targeted control measures before case numbers spike.
• Climate Adaptation: The findings challenge the assumption that Southern Brazil is unsuitable for year-round dengue transmission. The data suggests that Ae. aegypti is adapting to lower temperatures, sustaining transmission cycles even in autumn/winter.
• Policy Implications: The results emphasize that rainfall frequency is a more critical metric for vector control planning than total rainfall. This supports the need for continuous, rather than just seasonal, surveillance in temperate zones.
• Scalability: The modeling framework (LASSO with lagged variables) is presented as a generally applicable approach for predicting vector dynamics and disease risk in other regions facing climate-driven expansion of vector-borne diseases.

In conclusion, the paper provides robust evidence that climate variability is driving the expansion of dengue into Southern Brazil, with specific, quantifiable lags between environmental triggers and disease outcomes. This knowledge is vital for transitioning from reactive to proactive public health strategies in temperate urban environments.