Indian Ocean temperature anomalies predict long-term global dengue trends.

Despite identifying El Niño events as a factor in dengue dynamics, predicting the oscillation of global dengue epidemics remains challenging. Here, we investigate climate indicators and worldwide dengue incidence from 1990 to 2019 using climate-driven mechanistic models. We identify a distinct indicator, the Indian Ocean basin-wide (IOBW) index, as representing the regional average of sea surface temperature anomalies in the tropical Indian Ocean. IOBW is closely associated with dengue epidemics for both the Northern and Southern hemispheres. The ability of IOBW to predict dengue incidence likely arises as a result of its effect on local temperature anomalies through teleconnections. These findings indicate that the IOBW index can potentially enhance the lead time for dengue forecasts, leading to better-planned and more impactful outbreak responses.

Accelerating Progress Towards the 2030 Neglected Tropical Diseases Targets: How Can Quantitative Modeling Support Programmatic Decisions?

Over the past decade, considerable progress has been made in the control, elimination, and eradication of neglected tropical diseases (NTDs). Despite these advances, most NTD programs have recently experienced important setbacks; for example, NTD interventions were some of the most frequently and severely impacted by service disruptions due to the coronavirus disease 2019 (COVID-19) pandemic. Mathematical modeling can help inform selection of interventions to meet the targets set out in the NTD road map 2021-2030, and such studies should prioritize questions that are relevant for decision-makers, especially those designing, implementing, and evaluating national and subnational programs. In September 2022, the World Health Organization hosted a stakeholder meeting to identify such priority modeling questions across a range of NTDs and to consider how modeling could inform local decision making. Here, we summarize the outputs of the meeting, highlight common themes in the questions being asked, and discuss how quantitative modeling can support programmatic decisions that may accelerate progress towards the 2030 targets.

Chikungunya seroprevalence, force of infection, and prevalence of chronic disability after infection in endemic and epidemic settings: a systematic review, meta-analysis, and modelling study.

BACKGROUND: Chikungunya is an arboviral disease transmitted by Aedes aegypti and Aedes albopictus mosquitoes with a growing global burden linked to climate change and globalisation. We aimed to estimate chikungunya seroprevalence, force of infection (FOI), and prevalence of related chronic disability and hospital admissions in endemic and epidemic settings. METHODS: In this systematic review, meta-analysis, and modelling study, we searched PubMed, Ovid, and Web of Science for articles published from database inception until Sept 26, 2022, for prospective and retrospective cross-sectional studies that addressed serological chikungunya virus infection in any geographical region, age group, and population subgroup and for longitudinal prospective and retrospective cohort studies with data on chronic chikungunya or hospital admissions in people with chikungunya. We did a systematic review of studies on chikungunya seroprevalence and fitted catalytic models to each survey to estimate location-specific FOI (ie, the rate at which susceptible individuals acquire chikungunya infection). We performed a meta-analysis to estimate the proportion of symptomatic patients with laboratory-confirmed chikungunya who had chronic chikungunya or were admitted to hospital following infection. We used a random-effects model to assess the relationship between chronic sequelae and follow-up length using linear regression. The systematic review protocol is registered online on PROSPERO, CRD42022363102. FINDINGS: We identified 60 studies with data on seroprevalence and chronic chikungunya symptoms done across 76 locations in 38 countries, and classified 17 (22%) of 76 locations as endemic settings and 59 (78%) as epidemic settings. The global long-term median annual FOI was 0·007 (95% uncertainty interval [UI] 0·003-0·010) and varied from 0·0001 (0·00004-0·0002) to 0·113 (0·07-0·20). The highest estimated median seroprevalence at age 10 years was in south Asia (8·0% [95% UI 6·5-9·6]), followed by Latin America and the Caribbean (7·8% [4·9-14·6]), whereas median seroprevalence was lowest in the Middle East (1·0% [0·5-1·9]). We estimated that 51% (95% CI 45-58) of people with laboratory-confirmed symptomatic chikungunya had chronic disability after infection and 4% (3-5) were admitted to hospital following infection. INTERPRETATION: We inferred subnational heterogeneity in long-term average annual FOI and transmission dynamics and identified both endemic and epidemic settings across different countries. Brazil, Ethiopia, Malaysia, and India included both endemic and epidemic settings. Long-term average annual FOI was higher in epidemic settings than endemic settings. However, long-term cumulative incidence of chikungunya can be similar between large outbreaks in epidemic settings with a high FOI and endemic settings with a relatively low FOI. FUNDING: International Vaccine Institute.

Interactions between climate change, urban infrastructure and mobility are driving dengue emergence in Vietnam.

Dengue is expanding globally, but how dengue emergence is shaped locally by interactions between climatic and socio-environmental factors is not well understood. Here, we investigate the drivers of dengue incidence and emergence in Vietnam, through analysing 23 years of district-level case data spanning a period of significant socioeconomic change (1998-2020). We show that urban infrastructure factors (sanitation, water supply, long-term urban growth) predict local spatial patterns of dengue incidence, while human mobility is a more influential driver in subtropical northern regions than the endemic south. Temperature is the dominant factor shaping dengue's distribution and dynamics, and using long-term reanalysis temperature data we show that warming since 1950 has expanded transmission risk throughout Vietnam, and most strongly in current dengue emergence hotspots (e.g., southern central regions, Ha Noi). In contrast, effects of hydrometeorology are complex, multi-scalar and dependent on local context: risk increases under either short-term precipitation excess or long-term drought, but improvements in water supply mitigate drought-associated risks except under extreme conditions. Our findings challenge the assumption that dengue is an urban disease, instead suggesting that incidence peaks in transitional landscapes with intermediate infrastructure provision, and provide evidence that interactions between recent climate change and mobility are contributing to dengue's expansion throughout Vietnam.

A systematic review of the data, methods and environmental covariates used to map Aedes-borne arbovirus transmission risk.

BACKGROUND: Aedes (Stegomyia)-borne diseases are an expanding global threat, but gaps in surveillance make comprehensive and comparable risk assessments challenging. Geostatistical models combine data from multiple locations and use links with environmental and socioeconomic factors to make predictive risk maps. Here we systematically review past approaches to map risk for different Aedes-borne arboviruses from local to global scales, identifying differences and similarities in the data types, covariates, and modelling approaches used. METHODS: We searched on-line databases for predictive risk mapping studies for dengue, Zika, chikungunya, and yellow fever with no geographical or date restrictions. We included studies that needed to parameterise or fit their model to real-world epidemiological data and make predictions to new spatial locations of some measure of population-level risk of viral transmission (e.g. incidence, occurrence, suitability, etc.). RESULTS: We found a growing number of arbovirus risk mapping studies across all endemic regions and arboviral diseases, with a total of 176 papers published 2002-2022 with the largest increases shortly following major epidemics. Three dominant use cases emerged: (i) global maps to identify limits of transmission, estimate burden and assess impacts of future global change, (ii) regional models used to predict the spread of major epidemics between countries and (iii) national and sub-national models that use local datasets to better understand transmission dynamics to improve outbreak detection and response. Temperature and rainfall were the most popular choice of covariates (included in 50% and 40% of studies respectively) but variables such as human mobility are increasingly being included. Surprisingly, few studies (22%, 31/144) robustly tested combinations of covariates from different domains (e.g. climatic, sociodemographic, ecological, etc.) and only 49% of studies assessed predictive performance via out-of-sample validation procedures. CONCLUSIONS: Here we show that approaches to map risk for different arboviruses have diversified in response to changing use cases, epidemiology and data availability. We identify key differences in mapping approaches between different arboviral diseases, discuss future research needs and outline specific recommendations for future arbovirus mapping.

Projecting the future incidence and burden of dengue in Southeast Asia.

The recent global expansion of dengue has been facilitated by changes in urbanisation, mobility, and climate. In this work, we project future changes in dengue incidence and case burden to 2099 under the latest climate change scenarios. We fit a statistical model to province-level monthly dengue case counts from eight countries across Southeast Asia, one of the worst affected regions. We project that dengue incidence will peak this century before declining to lower levels with large variations between and within countries. Our findings reveal that northern Thailand and Cambodia will show the biggest decreases and equatorial areas will show the biggest increases. The impact of climate change will be counterbalanced by income growth, with population growth having the biggest influence on increasing burden. These findings can be used for formulating mitigation and adaptation interventions to reduce the immediate growing impact of dengue virus in the region.

Spatial dynamics of malaria transmission.

The Ross-Macdonald model has exerted enormous influence over the study of malaria transmission dynamics and control, but it lacked features to describe parasite dispersal, travel, and other important aspects of heterogeneous transmission. Here, we present a patch-based differential equation modeling framework that extends the Ross-Macdonald model with sufficient skill and complexity to support planning, monitoring and evaluation for Plasmodium falciparum malaria control. We designed a generic interface for building structured, spatial models of malaria transmission based on a new algorithm for mosquito blood feeding. We developed new algorithms to simulate adult mosquito demography, dispersal, and egg laying in response to resource availability. The core dynamical components describing mosquito ecology and malaria transmission were decomposed, redesigned and reassembled into a modular framework. Structural elements in the framework-human population strata, patches, and aquatic habitats-interact through a flexible design that facilitates construction of ensembles of models with scalable complexity to support robust analytics for malaria policy and adaptive malaria control. We propose updated definitions for the human biting rate and entomological inoculation rates. We present new formulas to describe parasite dispersal and spatial dynamics under steady state conditions, including the human biting rates, parasite dispersal, the "vectorial capacity matrix," a human transmitting capacity distribution matrix, and threshold conditions. An [Formula: see text] package that implements the framework, solves the differential equations, and computes spatial metrics for models developed in this framework has been developed. Development of the model and metrics have focused on malaria, but since the framework is modular, the same ideas and software can be applied to other mosquito-borne pathogen systems.

A phylogenetic study of dengue virus in urban Vietnam shows long-term persistence of endemic strains.

Dengue virus (DENV) causes repeated outbreaks of disease in endemic areas, with patterns of local transmission strongly influenced by seasonality, importation via human movement, immunity, and vector control efforts. An understanding of how each of these interacts to enable endemic transmission (continual circulation of local virus strains) is largely unknown. There are times of the year when no cases are reported, often for extended periods of time, perhaps wrongly implying the successful eradication of a local strain from that area. Individuals who presented at a clinic or hospital in four communes in Nha Trang, Vietnam, were initially tested for DENV antigen presence. Enrolled positive individuals then had their corresponding household members invited to participate, and those who enrolled were tested for DENV. The presence of viral nucleic acid in all samples was confirmed using quantitative polymerase chain reaction, and positive samples were then whole-genome sequenced using an amplicon and target enrichment library preparation techniques and Illumina MiSeq sequencing technology. Generated consensus genome sequences were then analysed using phylogenetic tree reconstruction to categorise sequences into clades with a common ancestor, enabling investigations of both viral clade persistence and introductions. Hypothetical introduction dates were additionally assessed using a molecular clock model that calculated the time to the most recent common ancestor (TMRCA). We obtained 511 DENV whole-genome sequences covering four serotypes and more than ten distinct viral clades. For five of these clades, we had sufficient data to show that the same viral lineage persisted for at least several months. We noted that some clades persisted longer than others during the sampling time, and by comparison with other published sequences from elsewhere in Vietnam and around the world, we saw that at least two different viral lineages were introduced into the population during the study period (April 2017-2019). Next, by inferring the TMRCA from the construction of molecular clock phylogenies, we predicted that two of the viral lineages had been present in the study population for over a decade. We observed five viral lineages co-circulating in Nha Trang from three DENV serotypes, with two likely to have remained as uninterrupted transmission chains for a decade. This suggests clade cryptic persistence in the area, even during periods of low reported incidence.

Combining rapid diagnostic tests to estimate primary and post-primary dengue immune status at the point of care.

BACKGROUND: Characterising dengue virus (DENV) infection history at the point of care is challenging as it relies on intensive laboratory techniques. We investigated how combining different rapid diagnostic tests (RDTs) can be used to accurately determine the primary and post-primary DENV immune status of reporting patients during diagnosis. METHODS AND FINDINGS: Serum from cross-sectional surveys of acute suspected dengue patients in Indonesia (N:200) and Vietnam (N: 1,217) were assayed using dengue laboratory assays and RDTs. Using logistic regression modelling, we determined the probability of being DENV NS1, IgM and IgG RDT positive according to corresponding laboratory viremia, IgM and IgG ELISA metrics. Laboratory test thresholds for RDT positivity/negativity were calculated using Youden's J index and were utilized to estimate the RDT outcomes in patients from the Philippines, where only data for viremia, IgM and IgG were available (N:28,326). Lastly, the probabilities of being primary or post-primary according to every outcome using all RDTs, by day of fever, were calculated. Combining NS1, IgM and IgG RDTs captured 94.6% (52/55) and 95.4% (104/109) of laboratory-confirmed primary and post-primary DENV cases, respectively, during the first 5 days of fever. Laboratory test predicted, and actual, RDT outcomes had high agreement (79.5% (159/200)). Among patients from the Philippines, different combinations of estimated RDT outcomes were indicative of post-primary and primary immune status. Overall, IgG RDT positive results were confirmatory of post-primary infections. In contrast, IgG RDT negative results were suggestive of both primary and post-primary infections on days 1-2 of fever, yet were confirmatory of primary infections on days 3-5 of fever. CONCLUSION: We demonstrate how the primary and post-primary DENV immune status of reporting patients can be estimated at the point of care by combining NS1, IgM and IgG RDTs and considering the days since symptoms onset. This framework has the potential to strengthen surveillance operations and dengue prognosis, particularly in low resource settings.

The local burden of disease during the first wave of the COVID-19 epidemic in England: estimation using different data sources from changing surveillance practices.

BACKGROUND: The COVID-19 epidemic has differentially impacted communities across England, with regional variation in rates of confirmed cases, hospitalisations and deaths. Measurement of this burden changed substantially over the first months, as surveillance was expanded to accommodate the escalating epidemic. Laboratory confirmation was initially restricted to clinical need ("pillar 1") before expanding to community-wide symptomatics ("pillar 2"). This study aimed to ascertain whether inconsistent measurement of case data resulting from varying testing coverage could be reconciled by drawing inference from COVID-19-related deaths. METHODS: We fit a Bayesian spatio-temporal model to weekly COVID-19-related deaths per local authority (LTLA) throughout the first wave (1 January 2020-30 June 2020), adjusting for the local epidemic timing and the age, deprivation and ethnic composition of its population. We combined predictions from this model with case data under community-wide, symptomatic testing and infection prevalence estimates from the ONS infection survey, to infer the likely trajectory of infections implied by the deaths in each LTLA. RESULTS: A model including temporally- and spatially-correlated random effects was found to best accommodate the observed variation in COVID-19-related deaths, after accounting for local population characteristics. Predicted case counts under community-wide symptomatic testing suggest a total of 275,000-420,000 cases over the first wave - a median of over 100,000 additional to the total confirmed in practice under varying testing coverage. This translates to a peak incidence of around 200,000 total infections per week across England. The extent to which estimated total infections are reflected in confirmed case counts was found to vary substantially across LTLAs, ranging from 7% in Leicester to 96% in Gloucester with a median of 23%. CONCLUSIONS: Limitations in testing capacity biased the observed trajectory of COVID-19 infections throughout the first wave. Basing inference on COVID-19-related mortality and higher-coverage testing later in the time period, we could explore the extent of this bias more explicitly. Evidence points towards substantial under-representation of initial growth and peak magnitude of infections nationally, to which different parts of the country contribute unequally.

Geo-Spatial Characteristics of 567 Places of Tick-Borne Encephalitis Infection in Southern Germany, 2018-2020.

Tick-borne encephalitis (TBE) is a growing public health problem with increasing incidence and expanding risk areas. Improved prevention requires better understanding of the spatial distribution and ecological determinants of TBE transmission. However, a TBE risk map at sub-district level is still missing for Germany. We investigated the distribution and geo-spatial characteristics of 567 self-reported places of probable TBE infection (POI) from 359 cases notified in 2018-2020 in the study area of Bavaria and Baden-Wuerttemberg, compared to 41 confirmed TBE foci and 1701 random comparator places. We built an ecological niche model to interpolate TBE risk to the entire study area. POI were distributed heterogeneously at sub-district level, as predicted probabilities varied markedly across regions (range 0-93%). POI were spatially associated with abiotic, biotic, and anthropogenic geo-spatial characteristics, including summer precipitation, population density, and annual frost days. The model performed with 69% sensitivity and 63% specificity at an optimised probability threshold (0.28) and an area under the curve of 0.73. We observed high predictive probabilities in small-scale areas, consistent with the known circulation of the TBE virus in spatially restricted microfoci. Supported by further field work, our findings may help identify new TBE foci. Our fine-grained risk map could supplement targeted prevention in risk areas.

Measuring the effects of COVID-19-related disruption on dengue transmission in southeast Asia and Latin America: a statistical modelling study.

BACKGROUND: The COVID-19 pandemic has resulted in unprecedented disruption to society, which indirectly affects infectious disease dynamics. We aimed to assess the effects of COVID-19-related disruption on dengue, a major expanding acute public health threat, in southeast Asia and Latin America. METHODS: We assembled data on monthly dengue incidence from WHO weekly reports, climatic data from ERA5, and population variables from WorldPop for 23 countries between January, 2014 and December, 2019 and fit a Bayesian regression model to explain and predict seasonal and multi-year dengue cycles. We compared model predictions with reported dengue data January to December, 2020, and assessed if deviations from projected incidence since March, 2020 are associated with specific public health and social measures (from the Oxford Coronavirus Government Response Tracer database) or human movement behaviours (as measured by Google mobility reports). FINDINGS: We found a consistent, prolonged decline in dengue incidence across many dengue-endemic regions that began in March, 2020 (2·28 million cases in 2020 vs 4·08 million cases in 2019; a 44·1% decrease). We found a strong association between COVID-19-related disruption (as measured independently by public health and social measures and human movement behaviours) and reduced dengue risk, even after taking into account other drivers of dengue cycles including climatic and host immunity (relative risk 0·01-0·17, p<0·01). Measures related to the closure of schools and reduced time spent in non-residential areas had the strongest evidence of association with reduced dengue risk, but high collinearity between covariates made specific attribution challenging. Overall, we estimate that 0·72 million (95% CI 0·12-1·47) fewer dengue cases occurred in 2020 potentially attributable to COVID-19-related disruption. INTERPRETATION: In most countries, COVID-19-related disruption led to historically low dengue incidence in 2020. Continuous monitoring of dengue incidence as COVID-19-related restrictions are relaxed will be important and could give new insights into transmission processes and intervention options. FUNDING: National Key Research and Development Program of China and the Medical Research Council.

Mapping environmental suitability of Haemagogus and Sabethes spp. mosquitoes to understand sylvatic transmission risk of yellow fever virus in Brazil.

BACKGROUND: Yellow fever (YF) is an arboviral disease which is endemic to Brazil due to a sylvatic transmission cycle maintained by infected mosquito vectors, non-human primate (NHP) hosts, and humans. Despite the existence of an effective vaccine, recent sporadic YF epidemics have underscored concerns about sylvatic vector surveillance, as very little is known about their spatial distribution. Here, we model and map the environmental suitability of YF's main vectors in Brazil, Haemagogus spp. and Sabethes spp., and use human population and NHP data to identify locations prone to transmission and spillover risk. METHODOLOGY/PRINCIPAL FINDINGS: We compiled a comprehensive set of occurrence records on Hg. janthinomys, Hg. leucocelaenus, and Sabethes spp. from 1991-2019 using primary and secondary data sources. Linking these data with selected environmental and land-cover variables, we adopted a stacked regression ensemble modelling approach (elastic-net regularized GLM, extreme gradient boosted regression trees, and random forest) to predict the environmental suitability of these species across Brazil at a 1 km x 1 km resolution. We show that while suitability for each species varies spatially, high suitability for all species was predicted in the Southeastern region where recent outbreaks have occurred. By integrating data on NHP host reservoirs and human populations, our risk maps further highlight municipalities within the region that are prone to transmission and spillover. CONCLUSIONS/SIGNIFICANCE: Our maps of sylvatic vector suitability can help elucidate potential locations of sylvatic reservoirs and be used as a tool to help mitigate risk of future YF outbreaks and assist in vector surveillance. Furthermore, at-risk regions identified from our work could help disease control and elucidate gaps in vaccination coverage and NHP host surveillance.

Tracking the emergence of disparities in the subnational spread of COVID-19 in Brazil using an online application for real-time data visualisation: A longitudinal analysis.

BACKGROUND: Brazil is one of the countries worst affected by the COVID-19 pandemic with over 20 million cases and 557,000 deaths reported by August 2021. Comparison of real-time local COVID-19 data between areas is essential for understanding transmission, measuring the effects of interventions, and predicting the course of the epidemic, but are often challenging due to different population sizes and structures. METHODS: We describe the development of a new app for the real-time visualisation of COVID-19 data in Brazil at the municipality level. In the CLIC-Brazil app, daily updates of case and death data are downloaded, age standardised and used to estimate the effective reproduction number (Rt ). We show how such platforms can perform real-time regression analyses to identify factors associated with the rate of initial spread and early reproduction number. We also use survival methods to predict the likelihood of occurrence of a new peak of COVID-19 incidence. FINDINGS: After an initial introduction in São Paulo and Rio de Janeiro states in early March 2020, the epidemic spread to northern states and then to highly populated coastal regions and the Central-West. Municipalities with higher metrics of social development experienced earlier arrival of COVID-19 (decrease of 11·1 days [95% CI:8.9,13.2] in the time to arrival for each 10% increase in the social development index). Differences in the initial epidemic intensity (mean Rt ) were largely driven by geographic location and the date of local onset. INTERPRETATION: This study demonstrates that platforms that monitor, standardise and analyse the epidemiological data at a local level can give useful real-time insights into outbreak dynamics that can be used to better adapt responses to the current and future pandemics. FUNDING: This project was supported by a Medical Research Council UK (MRC-UK) -São Paulo Research Foundation (FAPESP) CADDE partnership award (MR/S0195/1 and FAPESP 18/14389-0).

The importance of saturating density dependence for population-level predictions of SARS-CoV-2 resurgence compared with density-independent or linearly density-dependent models, England, 23 March to 31 July 2020.

BackgroundPopulation-level mathematical models of outbreaks typically assume that disease transmission is not impacted by population density ('frequency-dependent') or that it increases linearly with density ('density-dependent').AimWe sought evidence for the role of population density in SARS-CoV-2 transmission.MethodsUsing COVID-19-associated mortality data from England, we fitted multiple functional forms linking density with transmission. We projected forwards beyond lockdown to ascertain the consequences of different functional forms on infection resurgence.ResultsCOVID-19-associated mortality data from England show evidence of increasing with population density until a saturating level, after adjusting for local age distribution, deprivation, proportion of ethnic minority population and proportion of key workers among the working population. Projections from a mathematical model that accounts for this observation deviate markedly from the current status quo for SARS-CoV-2 models which either assume linearity between density and transmission (30% of models) or no relationship at all (70%). Respectively, these classical model structures over- and underestimate the delay in infection resurgence following the release of lockdown.ConclusionIdentifying saturation points for given populations and including transmission terms that account for this feature will improve model accuracy and utility for the current and future pandemics.

Assessing the impact of COVID-19 border restrictions on dengue transmission in Yunnan Province, China: an observational epidemiological and phylogenetic analysis.

BACKGROUND: In response to the COVID-19 pandemic, China implemented strict restrictions on cross-border travel to prevent disease importation. Yunnan, a Chinese province that borders dengue-endemic countries in Southeast Asia, experienced unprecedented reduction in dengue, from 6840 recorded cases in 2019 to 260 in 2020. METHODS: Using a combination of epidemiological and virus genomic data, collected from 2013 to 2020 in Yunnan and neighbouring countries, we conduct a series of analyses to characterise the role of virus importation in driving dengue dynamics in Yunnan and assess the association between recent international travel restrictions and the decline in dengue reported in Yunnan in 2020. FINDINGS: We find strong evidence that dengue incidence between 2013-2019 in Yunnan was closely linked with international importation of cases. A 0-2 month lag in incidence not explained by seasonal differences, absence of local transmission in the winter, effective reproductive numbers < 1 (as estimated independently using genetic data) and diverse cosmopolitan dengue virus phylogenies all suggest dengue is non-endemic in Yunnan. Using a multivariate statistical model we show that the substantial decline in dengue incidence observed in Yunnan in 2020 but not in neighbouring countries is closely associated with the timing of international travel restrictions, even after accounting for other environmental drivers of dengue incidence. INTERPRETATION: We conclude that Yunnan is a regional sink for DENV lineage movement and that border restrictions may have substantially reduced dengue burden in 2020, potentially averting thousands of cases. Targeted testing and surveillance of travelers returning from high-risk areas could help to inform public health strategies to minimise or even eliminate dengue outbreaks in non-endemic settings like southern China. FUNDING: Funding for this study was provided by National Key Research and Development Program of China, Beijing Science and Technology Planning Project (Z201100005420010); Beijing Natural Science Foundation (JQ18025); Beijing Advanced Innovation Program for Land Surface Science; National Natural Science Foundation of China (82073616); Young Elite Scientist Sponsorship Program by CAST (YESS) (2018QNRC001); H.T., O.P.G. and M.U.G.K. acknowledge support from the Oxford Martin School. O.J.B was supported by a Wellcome Trust Sir Henry Wellcome Fellowship (206471/Z/17/Z). Chinese translation of the abstract (Appendix 2).