Background

Future distribution of dengue risk is usually predicted based on predicted climate changes using general circulation models (GCMs). However, it is difficult to validate the GCM results and assess the uncertainty of the predictions. The observed changes in climate may be very different from the GCM results. We aim to utilize trends in observed climate dynamics to predict future risks of Aedes albopictus in China.

Methods

We collected Ae. albopictus surveillance data and observed climate records from 80 meteorological stations from 1970 to 2021. We analyzed the trends in climate change in China and made predictions on future climate for the years 2050 and 2080 based on trend analyses. We analyzed the relationship between climatic variables and the prevalence of Ae. albopictus in different months/seasons. We built a classification tree model (based on the average of 999 runs of classification and regression tree analyses) to predict the monthly/seasonal Ae. albopictus distribution based on the average climate from 1970 to 2000 and assessed the contributions of different climatic variables to the Ae. albopictus distribution. Using these models, we projected the future distributions of Ae. albopictus for 2050 and 2080.

Results

The study included Ae. albopictus surveillance from 259 sites in China found that winter to early spring (November-February) temperatures were strongly correlated with Ae. albopictus prevalence (prediction accuracy ranges 93.0-98.8%)-the higher the temperature the higher the prevalence, while precipitation in summer (June-September) was important predictor for Ae. albopictus prevalence. The machine learning tree models predicted the current prevalence of Ae. albopictus with high levels of agreement (accuracy > 90% and Kappa agreement > 80% for all 12 months). Overall, winter temperature contributed the most to Ae. albopictus distribution, followed by summer precipitation. An increase in temperature was observed from 1970 to 2021 in most places in China, and annual change rates varied substantially from -0.22 ºC/year to 0.58 ºC/year among sites, with the largest increase in temperature occurring from February to April (an annual increase of 1.4-4.7 ºC in monthly mean, 0.6-4.0 ºC in monthly minimum, and 1.3-4.3 ºC in monthly maximum temperature) and the smallest in November and December. Temperature increases were lower in the tropics/subtropics (1.5-2.3 ºC from February-April) compared to the high-latitude areas (2.6-4.6 ºC from February-April). The projected temperatures in 2050 and 2080 by this study were approximately 1-1.5 °C higher than those projected by GCMs. The estimated current Ae. albopictus risk distribution had a northern boundary of north-central China and the southern edge of northeastern China, with a risk period of June-September. The projected future Ae. albopictus risks in 2050 and 2080 cover nearly all of China, with an expanded risk period of April-October. The current at-risk population was estimated to be 960 million and the future at-risk population was projected to be 1.2 billion.

Conclusions

The magnitude of climate change in China is likely to surpass GCM predictions. Future dengue risks will expand to cover nearly all of China if current climate trends continue.

Abstract: Background. Future distribution of dengue risk is usually predicted based on predicted climate changes using general circulation models (GCMs). However, it is hard to validate the GCM results and assess the uncertainty of the predictions. Observed changes in climate may be very different from GCM results. Methods.We collected Aedes albopictus surveillance data and observed climate records from about 90 meteorological stations for the period 1970–2021. We analyzed the trends of climate change in China and made predictions on future climate for the years 2050 and 2080 based on trend analyses. We analyzed the relationship between climatic variables and the prevalence of Ae. albopictusin different months/seasons. We built a series of machine learning classification tree models to predict the monthly/seasonal Ae. albopictus distribution based on the average climate from 1970 to 2000 and assessed the contributions of different climatic variables to the Ae. albopictus distribution. Using these models, we projected the future distributions of Ae. albopictus for the years of 2050 and 2080. Results. The study included Ae. albopictus surveillance from 259 sites in China and found that winter to early spring (November–February) temperatures were strongly correlated with Ae. albopictus prevalence – the higher the temperature the higher the prevalence, while precipitation in summer (June–September) was important predictors for Ae. albopictus prevalence. The machine learning tree models predicted the current prevalence of Ae. albopictus with high agreement (accuracy > 90% and Kappa agreement > 80% for all months). Overall, winter temperature contributed the most to Ae. albopictus distribution, followed by summer precipitation. Increase in temperature was observed in most places in China and rates of annual change varied substantially among sites, with the highest increase in temperature occurring from February to April (annual increase of 1.4 – 4.7ºC for monthly mean, 0.6 – 4.0ºC for monthly minimum, and 1.3 – 4.3ºC for monthly maximum temperature) and the lowest in November and December. Temperature increases were lower in the tropics/subtropics compared to the high-latitude areas. The projected temperatures in 2050 and 2080 by this study were about 1~1.5°C higher than projected by GCMs. The estimated current Ae. albopictus risk distribution had a northern boundary of north-central China and the southern edge of northeastern China, with a risk period of June–September. The projected future Ae. albopictus risks in 2050 and 2080 cover nearly all of China, with an expanded risk period of April–October. The current at-risk population was estimated to be 960 million and the future at-risk population was projected to be 1.2 billion. Conclusions.The magnitude of climate change in China is likely to surpass the GCM predictions. Future dengue risks will expand to cover nearly all of China if the current climate trends continue.

Document recommendation systems for locating relevant literature have mostly relied on methods developed a decade ago. This is largely due to the lack of a large offline gold-standard benchmark of relevant documents that cover a variety of research fields such that newly developed literature search techniques can be compared, improved and translated into practice. To overcome this bottleneck, we have established the RElevant LIterature SearcH consortium consisting of more than 1500 scientists from 84 countries, who have collectively annotated the relevance of over 180 000 PubMed-listed articles with regard to their respective seed (input) article/s. The majority of annotations were contributed by highly experienced, original authors of the seed articles. The collected data cover 76% of all unique PubMed Medical Subject Headings descriptors. No systematic biases were observed across different experience levels, research fields or time spent on annotations. More importantly, annotations of the same document pairs contributed by different scientists were highly concordant. We further show that the three representative baseline methods used to generate recommended articles for evaluation (Okapi Best Matching 25, Term Frequency-Inverse Document Frequency and PubMed Related Articles) had similar overall performances. Additionally, we found that these methods each tend to produce distinct collections of recommended articles, suggesting that a hybrid method may be required to completely capture all relevant articles. The established database server located at https://relishdb.ict.griffith.edu.au is freely available for the downloading of annotation data and the blind testing of new methods. We expect that this benchmark will be useful for stimulating the development of new powerful techniques for title and title/abstract-based search engines for relevant articles in biomedical research.