Introduction

Human beings are at the interface of the earth’s physical, geological and biological processes. Any fundamental global environmental change affecting these systems will drastically impact human health in various ways. The World Health Organization (WHO) claims that global warming that occurred since the 1970s had caused over 140, 000 deaths annually by the year 2004, proclaimed to be significant evidence for climate change’s impact on human health. According to the World Malaria Report 2021, 241 million malaria cases and 627000 malaria deaths occurred worldwide in 2020, mainly attributed to the disruptions caused due to the COVID-19 pandemic.

In this connection, climate change plays an important role in defining various aspects of the malaria biology cycle. It has been pointed out by leading organizations like the Centre for Disease Control and Prevention (CDC) that climate is a key determinant in the geographic distribution and the seasonality of malaria. In general, temperature and rainfall are the two key climatic parameters which influence malaria transmission rate significantly. For instance, rainfall can create breeding sites where Anopheles mosquito can complete the aquatic life stages from egg to larvae to pupae which finally develop into its adulthood stage. The aquatic life stages of the Anopheles are completed in approximately 9–12 days in tropical areas1. At certain situation breeding sites may dry up impulsively in the absence of further rainfall or flushing of breeding sites might occur and destroy the Anopheles eggs or larvae by excessive rains2,Paaijjmans et al.,3) and the survival of an adult mosquito and its propagation is highly dependent on ambient temperature, humidity, and rainfall.

The impact of climate variables on malaria transmission is widely studied worldwide and in India as well. According to the recent reports, climate change plays a major role for the present malaria disease’s spread in India4,5,6,7. Temperature and rainfall play a major role in the development of the many stages of a mosquito’s life cycle, including blood-feeding rate, biting rate, gonotrophic cycle etc.8,9,10. Especially, rainfall over India lowers the country’s temperature and makes water available for mosquito breeding and their resting grounds, especially during the summer monsoon season.

The recent Intergovernmental Panel on Climate Change (IPCC) report suggests that the global average temperatures over the entire land and ocean surfaces have warmed roughly by 1 °C since the pre-industrial levels with more emphasis on the potential effect of climate change on human health in the future. The ongoing special report of IPCC on the impacts of global warming of 1.5 °C above pre-industrial levels noted with high confidence that any increase in global temperature (e.g., + 0.5 °C) shall affect human health, with primarily negative consequences (IPCC11).Most of the global climate models within the CMIP6 indicate a consistent increase in monsoon rainfall in India under global warming and there would be an annual mean surface temperature rise by end of the century ranging from 3 to 5 °C under different emission scenarios12, IPCC, 2022,13. In the warming climate, the intensity and frequency of extreme rainfall events in India are also predicted to increase, with the northeast and southern regions experiencing such stronger events. Findings from the previous studies on future climate change over Indian region suggests that due to global warming, the summer monsoon months (June to September) will receive higher rainfall in the future over most parts of India, mainly over the monsoon core zone under different IPCC scenarios14, Jayshankar et al.15,16,17,18,19,20. To evaluate the impact of anticipated climate change on malaria, several mathematical and numerical models have attempted to relate all the variables that affect malaria transmission. Some of those models relate variables related to climate change such as temperature and rainfall with some of the non-climatic factors such as human population, elevation and parameters related to mosquito life cycle (Martens et al. 1995). Many studies were conducted to understand the intricate relationship between climate and malaria. Further, Grasso et al.21 reviewed various quantitative methods used in modelling climate change impacts on human health with special case study on malaria. These included the time series models, cross-section and panel data studies, non-statistical and general equilibrium studies. These studies have been carried out in different geographical extent4,7,22,23,24,25,26. All these studies indicated that climate parameters have a major role in changing the spatio-temporal distribution of malaria. It is evident that there are a range of models available which are used to predict the future malaria transmission dynamics with respect to the climate change impacts. Both the statistical models and the biological models are used to predict the spatial and temporal distribution of malaria across the globe on a regional scale27, Beguin et al.28, van Leishout29,30,31. For instance, model simulations for future scenarios in Britain suggested that the change in temperature projected to occur by 2050 would likely cause a proportional increase in local malaria transmission of 8–14% (Kunh et al.32). A work by Caminade et al. (2014) adopted five malaria models driven by climate outputs from five global climate models (GCMs) using the four Representative Concentration Pathways (RCPs) emissions scenarios developed for the 5th IPCC assessment report are used to produce a new set of health global impact assessments. The multi-malaria models used for the purpose were the Liverpool Malaria Model (LMM), the VECTRI malaria model, the MARA malaria model, Umea University’s statistical model and Modeling Framework for the Health Impact Assessment of Man-Induced Atmospheric Changes (MIASMA) model. The results of the multi-malaria model, multi-GCM, multi-scenario inter comparison exercise were found to be consistent with previous studies in indicating that the most significant climate change effects were confined to the highlands in Africa and parts of South America and Southeast Asia.

Result from few modelling studies suggest that future projections of populations are at risk for malaria disease owing to climate change in few parts of the world but a large decrease in malaria endemicity has been observed worldwide, mainly owing to human intervention33. Malaria has been significantly reduced in India19,34,35. In this aspect, assessing the impact of climate variability on the dynamics of malaria transmission remains a significant challenge in the physio graphically diverse regions of India. The region has seen only a few studies utilizing models to forecast malaria transmission dynamics concerning both climate-related factors (such as temperature and rainfall) and other non-climate elements (including malaria parasites, vectors, human host attributes, and population density) that contribute to malaria transmission. For instance, Chaturvedi and Dwivedi19 quantified the spatio-temporal malaria transmission intensity in India using VECTRI, investigating the effects of climate change on malaria transmission variability from past (1961–2005) to future (2006–2050) periods. They have used global data sets from CMIP5 global climate models under different RCP scenarios. VECTRI output corresponding to 12 models are averaged for each projection scenarios (RCP2.6, RCP4.5, RCP8.5) to create the multi model mean (MMM) data and the results show that there is a dry bias observed in the MMM model during the monsoon season (JJA) when compared to the IMD observation. Whereas in this current study, results from CMIP6 simulation show that rainfall bias for the monsoon season is improved with a lesser bias when compared to the CMIP5 simulations of previous studies. Considering the maximum growth of vector density during the monsoon season, we hope that CMIP6 data would be a better choice for the VECTRI model to be used as inputs that enhances the reliability of malaria transmission dynamics in India. In another study, Parihar et al.35 employed VECTRI to study potential future malaria transmission in Odisha, considering CMIP5 models under RCP scenarios. This study provide insights into future malaria transmission in an malaria endemic region in India i.e. the state of Odisha, situated in the eastern parts of India. This research is particularly noteworthy in Odisha that establishes connections between climate variability and malaria transmission across its diverse districts. In response to the need for enhanced understanding of malaria transmission dynamics over different climate zones of India , our study extends further by using latest IPCC CMIP6 global model data sets to predict the future malaria transmission dynamics over different climate zones of India on a annual, seasonal and monthly scales. Consequently, the primary aim of this present research is to examine the fluctuations in malaria transmission patterns across four major regions of India through the utilization of the VECTRI model, which is influenced by a range of climatic and non-climatic elements. The existing model parameters incorporate the behaviors of the Anopheles gambiae vector and the Plasmodium falciparum parasite, observed within African contexts. Nonetheless, it is noteworthy that the Anopheles gambiae vector does not exist in the study area; rather, the focus is on the Anopheles culicifacies species.

Few of the past studies suggest that an culicifacies has a wide distribution in India which has been regarded as an predominant vector of malaria generating maximum malaria cases (above 60% annually) in the country36,Subbarao37, Sharma & Dev38). In particular, this species is one of the potent species in central India that accounts for 75% of the total mosquitoes species which prevails mainly during July to September, during rainy season (Nagpal & Sharma39, Subbarao40,41 and the transmission season is generally from May to November (Neeru et al.42). In general, An. culicifacies populations peak in monsoon and post-monsoon months. This particular species appears mostly in April and May and increases maximum in June and July and then starts declining in September (Asima, 2015). In particular, A. culicifacies shows strong seasonality aligned with monsoon patterns in India, with peak vector densities usually occurs about one month after peak rainfall43,44,Further, this mosquito species thrives well in plains receiving fair amount of rainfall. This also plays a major role in occurence of malaria transmission in many forested areas as well as in the hilly areas of Indian regions (Mathur et al., 1990; Bhuyan et al.45,46). A. culicifacies is behaviorally different from A. gambiae, and we acknowledge this limitation. A. culicifacies typically breeds in sunlit, freshwater pools, irrigation channels, and rice fields37,47, whereas A. gambiae prefers small, temporary, sunlit water bodies. Unlike A. gambiae, A. culicifacies is predominantly endophilic (rests indoors) but can exhibit exophilic behavior in certain conditions48. We accounted for this while changing the hydrological parameters in the VECTRI model run. Though we recognize that a complete reparameterization of all the vector dynamics would be ideal, but we adopted a targeted approach for the following reasons:

Firstly, considering the fact that An. culicifacies differs from An. gambiae in certain behavioral aspects, the fundamental biological responses to temperature and rainfall that drive the aquatic stages of mosquito development49,50 are being considered the major factors to compare the dynamics of the two species, justifying our focus on modifying these critical parameters. Secondly, the vector density of An. culicifacies is strongly correlated to temperature and rainfall, which played a major role in the rising of the vector prevalence in India51. In this context, since our primary objective is to assess climate change impacts on malaria transmission dynamics in India, we have focussed mainly on these two important climate-sensitive parameters such as temperature and rainfall and modified different parameters in the VECTRI model those are directly linked to the malaria transmission dynamics.

We have incorporated the biological processes and behaviors of Anopheles culicifacies species (details are given in the methodology section) and by adjusting the model’s parameterization scheme, we have conducted simulations across four regions of India for the baseline period of 30 years (1985–2014) and short-, medium- and long-term future projections under IPCC CMIP6 SSP5-8.5 scenario which is a business-as-usual scenario to evaluate the dynamics of malaria transmission. Pertinent variables like temperature, rainfall, and population density have been supplied to facilitate the operation of the VECTRI model. Additionally, a series of parameters have been modified to ensure the generation of dependable and realistic evaluations of parameters associated with the malaria vector and the progression of malaria transmission. The present study focusses on understanding the effect of climate change in 17 states of India, with distinct physio-geographical features and climatic conditions.

Study area

Considering the temporal and spatial distributions of malaria transmission, temperature and rainfall patterns during different monsoon seasons in India, we have classified the whole India into four major malaria prone zones: Central, Peninsular, Western, and Eastern zones. The four major malaria prone zones of India with distribution of 17 states is as follows and is depicted in Fig. 1:

  • Western: Maharashtra, Gujarat, Rajasthan, Goa

  • Central: Bihar, Madhya Pradesh, Uttar Pradesh

  • Eastern: West Bengal, Odisha, Chhattisgarh and Jharkhand

  • Peninsular: Andhra Pradesh, Karnataka, Telangana, Tamil Nadu, Kerala,

Fig. 1
figure 1

Study area map, (GIS software (version 10.2.1) is used for plotting this figure: Url link:https://support.esri.com/en-us/patches-updates/2017/arcgis-10-2-1-for-desktop-support-for-windows-10-patch-7462).

Full size image

Western region

(Approx. 15.0° to 24.0° N lat and 68.0° to 77.5° E lon): Western region is bordered by the Arabian Sea to the west, providing it with a long coastline that stretches from the Gujarat state in the north to the state of Maharashtra in the south. This coastline has contributed to trade, maritime activities, and cultural exchanges over centuries. Several important rivers flow through Western India, including the Sabarmati, Narmada, Tapi, and Godavari. These rivers are vital for agriculture, irrigation, and hydroelectric power generation. The southwestern monsoon, which arrives from the Arabian Sea, significantly impacts the climate of Western India. The monsoon season typically lasts from June to September, bringing heavy rainfall to the region and contributing to the fertile soils. The Western Ghats receive some of the highest rainfall in the country due to the monsoon winds.

Central region

(Approx. 19.0° to 24.5° N Lat and 75.0° to 82.0° E Lon): Central region has hot and dry summers, as the plateau topography contributes to temperature variations. The region receives a significant portion of its annual rainfall during this time, with the southwest monsoon winds bringing moisture-laden air from the Arabian Sea. Due to its vastness and varying topography, Central India exhibits climatic variations across its different sub-regions.

Eastern region

(Approx. 20.0° to 27.5° N Lon): This region is characterized by diverse geographical features. The Gangetic plains are fertile and well-suited for agriculture, owing to the deposition of alluvial soil by the Ganges and other rivers. The Himalayan Foothills, Bengal Delta and Coastline along the Bay of Bengal, which extends from West Bengal to Odisha and Andhra Pradesh are the characteristic features of this region. Predominantly receives rainfall from the Tropical Monsoon Climate characterized by distinct wet and dry seasons.

(Approx. 8.0° to 21.0° N to 72.0° to 82.0° E Lon): Peninsular region is primarily characterized by its rough terrain, elevated plateaus, and rolling hills. It is bounded by the Arabian Sea to the west, the Bay of Bengal to the east, and the Indian Ocean to the south. The climate of Peninsular India varies widely. The western coastal areas experience a tropical maritime climate, while the interior regions have a more arid or semi-arid climate. The Western Ghats receive heavy rainfall, leading to the formation of lush rainforests and biodiversity hotspots.

Methodology

VECTRI model is used in this study to simulate the malaria dynamics in the form of malaria transmission during June to September, south west monsoon (hereafter referred as JJAS) and October to December, northeast monsoon (hereafter refereed as OND) seasons of India. The detailed methodological flow chart id given in Fig. 2. In this study, VECTRI model is run over certain endemic regions of India as mentioned earlier (western, central, eastern and peninsular) by using the input data as daily rainfall, temperature and human population density data to predict the dynamics of the malaria transmission of the Plasmodium falciparum parasite that is transmitted through the Anopheles culicifacies vector.

Fig. 2
figure 2

Flow diagram showing the steps taken during the VECTRI model simulations procedures over Indian region.

Full size image

From the previous studies, it is found that P. falciparum parasite is one of the major cause of malaria transmission in India and An. culicifacies is likely to be considered as one of the major malaria vectors in these regions, which is prevalent throughout the year (Ranjit, 2006; Dhangadamajhi et al. 2015). In India, Plasmodium falciparum accounts for 63.38% of malaria cases, while Plasmodium vivax accounts for about 33–34% (Directorate of National Vector Borne Disease Control Programme52. Anopheles culicifacies is responsible for about 65% of malaria transmission in India54,55,55. VECTRI model is a climate sensitive model which takes climate parameters as major inputs, and the biology of mosquito species is fully dependent on climatic factors, we have modified the threshold values of temperature and rainfall at which the mosquito species can survive. From the previous studies, it is well known that the optimum temperature which favors larval development is 27 °C56. if the temperature is low beyond a threshold, larvae and adult mosquito survivability rate is very low57,58. In particular, minimum temperature of 16.5 to 18 °C as the threshold temperature values required for the development of Plasmodium falciparum parasite (Martens et al. 1995). Therefore, while simulating the model, the parameterization scheme is being modified according to the various climatic factors such as temperature and rainfall those are responsible for the survival of this particular malaria vector An. culicifacies. Temperature and rainfall affects the sporogonic and gonotrophic cycle development rates of the vectors. Higher temperature increases the mortality rates for adult vectors59 and heavy rainfall may flush off the existing breeding sites and decrease early-stage larvae through flushing3. Rainfall also facilitates enhanced humidity range 60–80%, suitable for vector survivability and developmental suitability54,55. In this aspect, the parameters are modified in the model based on the biological processes of the Anopheles culicifacies vector and the P. falciparum parasite and the parameters are modified by referring few of the available literatures (Raghavendra et al. 2010, Sahu et al. 2014, Tompkins & Ermert, 2013,60).

Studies by Barik et al.48 and Bhatt et al.61 show that A. culicifacies in India exhibits blood-feeding patterns that align with the default parameter ranges in the VECTRI model. Considering this, in our simulation, we have set the default value of blood meal success rate. Further, An. culicifacies exhibits varying anthropophilic (human-biting) and zoophilic (animal-biting) behaviors across its sibling species. Most of its sibling species are primarily zoophilic and has low anthropophilic indices ranging from 1 to 4%51. It is found that A. culicifacies has a lower human blood index (Anthropophilic index is between 0.3 to 0.5) when compared to A. gambiae37,62. This zoophilic tendency is reflected in our simulation results, where the EIR values are appropriately calibrated to match observed transmission intensities in India. We have modified this parameter accordingly in our study while simulating the VECTRI model. Secondly, as demonstrated in our analysis of using the temperature and rainfall data sets both from CMIP6 baseline simulation data and IMD observation data for the period 1985–2014, the VECTRI model with our modified parameters successfully reproduces the seasonal patterns of malaria transmission over the different climatic zones of India, indicating that our parameter selection effectively captured the essential dynamics.

The list of parameters with the description and their values set up in the model are given in Table 1. Temperature and rainfall are not simulated by VECTRI and hence VECTRI does not predict their future trends. It basically reproduces the rainfall and temperature in the VECTRI model as inputs to predict the malaria transmission dynamics.

Table 1 List of parameters modified during the VECTRI model simulations for the Indian region.
Full size table

VECTRI uses human population density data as one of the major inputs while calculating biting rates. The human biting rate is the number of Anopheles vectors biting an individual over a fixed period of time (Warrell & Gilles,63,64. The intensity of malaria transmission is measured with the help of Entomological Inoculation rate (EIR), which is the number of infectious bites per person per unit time. It is calulated as the product of the human biting rate and the sporozoite rate. i.e. the product of number of Anopheles per person and the average number of persons bitten by one Anopheles in one day (Shaukat et al.65). Urban areas with high population density will have a lower EIR due to lower parasite ratios when compared to the rural areas34.

The observed daily mean, maximum and minimum temperatures of 1° × 1° and rainfall data of 0.25° × 0.25° horizontal resolutions for the 30 years observed period (1985–2014) are obtained from India Meteorological Department (IMD) for the selected study areas for the VECTRI model simulation. Again, daily temperature and rainfall data (bias corrected) from the MMM of different CMIP6 global climate models are obtained for the baseline period (1985–2014) as well as for three future projection periods 2030s (2015–2035), 2050s (2035–2065), and 2080s (2065–2095). The population data is obtained from https://sedac.ciesin. columbia.edu /data/set/ popdynamics- global-pop-density- time-series-estimates (CIESN, 2017). Initially, population data are downloaded in a decadal interval i.e. for the years 1970, 1980, 1990, 2000, 2010, 2020, and 2030, 2040, 2050, 2060, 2070, 2080 and 2090 at a resolution of 1 sq km. The grids are adjusted by using rates of population change between decades. These data do not represent census observations for the years prior to 2000, and therefore can at best be thought of as estimations of the populations in a given location. All the gridded datasets used in the present research have been regridded to 0.25 × 0.25 degree resolution and have given as inputs to the VECTRI model.

Rainfall (mm day− 1), mean temperature (°C), mosquito density or vector density (per square meter) and entomological inoculation rate (EIR) in number of infective bites per person per day (i/b/d) are the few parameters generated by the VECTRI model, considered for our analysis. VECTRI outputs generated from MMM-simulations and from IMD observed data during the period 1985–2014 are compared and the difference between VECTRI model outputs (bias), forced with IMD observation and then VECTRI model outputs, forced with bias corrected multi model mean data obtained from CMIP6 climate models for all the selected output parameters are estimated to evaluate the VECTRI model performances over India. Then the climate change assessments are done by estimating the future change in temperature and rainfall with respect to the baseline and the corresponding change in vector density and EIR have been estimated to estimate the future change in malaria tranmission over the study period.

In general, the analyses include model validation, monthly and seasonal distributions of temperature, rainfall, vector density and EIR for the future projections period 2030s, 2050s and 2080s to predict the potential future change in dynamics of the malaria transmission over the study area. Since malaria transmission is largely seasonal corresponding to rainy season and maximum number of malaria cases and deaths occur during the monsoon seasons in India44, we have focused future malaria transmission dynamics during JJAS and OND in our analyses.

Results and discussions

VECTRI model evaluation

Using the approach outlined in the preceding section, we conduct simulations of malaria transmission using the VECTRI model specifically focused on the Indian region. These simulations encompass historical conditions as well as perspectives related to climate change, employing corresponding rainfall and temperature datasets from the CMIP6 global climate models. A critical aspect in this context is evaluating the compatibility of the CMIP6 climate model-derived rainfall and temperature data (utilized as input for the VECTRI model) with the higher-resolution IMD rainfall and temperature data.

Our investigation dealt with a comparison of the seasonal patterns in CMIP6 multi-model mean rainfall and temperature for two monsoon seasons, JJAS and OND across the four selected regions in India during the period 1985–2014 against the corresponding IMD dataset. Figure 3a−p shows the bias and correlation between the VECTRI generated parameters forced with MMM data and IMD data for the JJAS season. While estimating the bias, it is observed that the spatial rainfall distribution and the intensity is captured well by the CMIP6 MMM over most of the regions with a dry bias of 2 to 4 mm/day except over peninsular region, where there is a wet bias of 1 to 2 mm/day is observed (Fig. 3a–d). Notably, during the south west monsoon period, most of these regions experiences the maximum rainfall and the model shows a good agreement with the observation.

Fig. 3
figure 3

Estimating bias and correlation for all the derived parameter. between VECTRI outputs forced with CMIP6 MMM data and VECTRI outputs forced with IMD observation data during JJAS season for the period 1985–2014 [NCAR Command Language (NCL version 6.6.2) is used to generate this figure].

Full size image

Further, spatial patterns of mean temperature obtained from the CMIP6 MMM shows good agreement with the IMD observed data across most of the Indian regions. Whereas, a warm bias of 1 to 2 ⁰C is seen over the western and central India and cold bias of 1 to 2⁰C over peninsular region. There is a negative correlation in temperature seen between the CMIP6 model and IMD observation over the central region, transcending through the northern part of the Peninsular Region Fig. 3 (e–h).

VECTRI model is forced with two major climate variables such as temperature and rainfall and these two variables are obtained from CMIP6 MMM data and also from IMD observation data and are given inputs to the VECTRI model simulations. VECTRI model generates two different outputs for vector density and EIR. One is generated from the CMIP6 model data and the other one is generated from the IMD observation data. Since there is no available observed data for vector density and EIR for the Indian region, we have estimated bias for these two parameters from the outputs generated by VECTRI simulations with two different forcing data. It is found that the vector density is overestimated during the JJAS period using the CMIP6 model data over most of the Indian regions except few places over western parts where it is underestimated with respect to the IMD observation data however both the data sets are strongly correlated with each other (Fig. 3 (i-l)). This over estimation is also reflected in the EIR where the VECTRI forced with CMIP6 model shows higher EIR than the VECTRI forced with IMD observation during the period 1985–2014 (Fig. 3m−p).

Figure 4a-p shows the bias and correlation between the VECTRI generated parameters forced with MMM data and IMD data for the OND season. In general, peninsular India receives maximum rainfall during the Northeast monsoon season i.e. from October to December. Especially, During the OND period, Tamil Nadu receives about 48% of the annual total rainfall16,17. Results indicate that the rainfall spread across the peninsular region is well captured by the CMIP6 MMM when compared to the IMD observation during the OND period. The rainfall bias (model-observation) between CMIP6 MMM and IMD observation is presented in Fig. 4(a-d). VECTRI simulation using MMM data shows a reasonable skill in simulating OND rainfall patterns for all the four regions as compared to the IMD data, though over peninsular India, the model underestimates 1 to 2 mm/day when compared to the observed rainfall. Also,rainfall data between the model and observed data shows a significant correlation (0.4).

Fig. 4
figure 4

Same as Fig. 3 but for OND season [NCAR Command Language (NCL version 6.6.2) is used to generate this figure].

Full size image

The mean temperature varies between the range of 17–25⁰C in both observation and model and the model underestimates about 1 to 2⁰C over most of the regions as ccompared to the observation. A positive correlation between the two data sets is observed across Western and Peninsular region as shown in Fig. 4(e–h). Further, mosquito density and the EIR are slightly overestimated in the model when compared to the observation. Also, in both the data sets , there is a significant correlation of 0.4 observed in most of the regions as shown in Fig. 4(i-p).

Future projections

On the basis that the MMM reasonably simulates the temperature and rainfall patterns well over the four regions of India for the baseline period as seen before, the rainfall and temperature datasets of the CMIP6 MMM is analysed for the future climate scenarios (2015–2100).This aids to understand how the temperature and rainfall will change in the future climate and what will be its likely impact on the future malaria transmission rate during JJAs and OND under IPCC SSP5-8.5 scenario, which is buisness as usual scenario. The trend of temperature, rainfall and the associated vector related parameters such as mosquito density and EIR for the JJAS and OND seasons are plotted in Figs. 5 and Fig. 6, respectively.

Fig. 5
figure 5

Future trend of change in temperature (red solid line) in °C, rainfall (blue solid line) in % and EIR (green solid line) in % simulated by VECTRI model forced with CMIP6 MMM data during JJAS season.[R software (version 3.6.2) is used for plotting this figure].

Full size image
Fig. 6
figure 6

Future trend of change in temperature (red solid line) in °C, rainfall (blue solid line) in % and EIR (green solid line) in % simulated by VECTRI model forced with CMIP6 MMM data during OND season.[R software (version 3.6.2) is used for plotting this figure].

Full size image

It is observed from Fig. 5 that in all the four regions, temperature and rainfall experienced a positive trend in the future during the JJAS and EIR shows a decreasing trend across all regions except over peninsular and western region, where the EIR shows an increasing trend by the end of the century. Increase in temperature and rainfall over Western and Peninsular regions are in good agreement with decrease in EIR and the trend continues to persist till the end of century.

Figure 6 shows future trends in temperature, rainfall and EIR for all the four different regions of India. Results indicate that there is no significant trend in rainfall is observed during the OND whereas temperature and the EIR show an increasing trend over all the regions for the future period. Especially, a closest association of the increasing EIR with the increasing rainfall is found over Peninsular region during this post monsoon season. It could also be inferred that the EIR in the Peninsular region is influenced by both the increase in temperature and rainfall that makes favourable conditions for the mosquito vectors to grow causing the spread of malaria transmission during both the JJAS and OND season.

The monthly mean values of VECTRI simulated parameters over western,central, eastern and peninsular regions for the historical and three projection periods (2030s, 2050s and 2080s) are shown in Fig. 7, Fig. 8, Fig. 9 and Fig. 10 ,respectively.

Fig. 7
figure 7

Monthly mean values of VECTRI simulated parameters over western region for the present as well as for the three projection periods (2030s, 2050s and 2080s)[R software (version 3.6.2) is used for plotting this figure].

Full size image
Fig. 8
figure 8

Same as Fig. 7 but for eastern region. [R software (version 3.6.2) is used for plotting this figure].

Full size image
Fig. 9
figure 9

Same as Fig. 7 but for central region [R software (version 3.6.2) is used for plotting this figure].

Full size image
Fig. 10
figure 10

Same as Fig. 7 but for peninsular region, [R software (version 3.6.2) is used for plotting this figure].

Full size image

In general, the data from CMIP6 global model data predicts an increase in the mean temperature (around 4 to 5 °C by end of the century, 2080s) and rainfall (around 20 to 50% by 2080s) during both the summer and winter monsoon seasons with reference to baseline period. In particular, an increase in the daily rainfall during the JJAS seen for all the projection periods i.e. for 2030s, 2050s and 2080s compared to the baseline period. From the VECTRI model simulation, the monthly distribution of malaria transmission in terms of EIR and vector density follows similar trend as seen in the rainfall patterns over all the regions. The peaks in the EIR and vector density are predominant in the August month during historical and future periods, which is a month later than the peak of rainfall in July. It is interesting to note that during summer monsoon and winter monsoon period of baseline period, the vector density and EIR values show higher, suggesting higher transmission rates of malaria in a monsoon environment whereas the mean monthly EIR shows a decrease or stabilization after 2050s when compared to the baseline period and the mean monthly variability of vector density for the projection periods 2030s, 2050s, and 2080s shows an increase. Similar patterns are observed across the eastern, peninsular, and western region.

Projections during JJAS.

Figure 11 (a-c) shows the percentage change in rainfall for the projection periods 2030s, 2050s, and 2080s compared to the baseline period during JJAS and the results indicate that future rainfall over all the four regions are likely to increase by 10 to 50% and more. maximum increase is observed in the entire Western region and peninsular region (above 50%) by 2080s with respect to the baseline period. Other regions show marginal increase in rainfall by 10–20% by end of the century.

Fig. 11
figure 11

Future Projections of all the VECTRI simulated four parameters during JJAS season.[NCAR Command Language (NCL version 6.6.2) is used to generate this figure].

Full size image

The mean temperature over all the four regions during JJAS baseline period shows about 25 to 30 °C. The future projections of mean temperature over most of the Indian regions show an increase of 0.5 to 1 °C, 1 to 2 °C and 2.5 to 3.5 °C for the periods 2030s, 2050s and 2080s, respectively, with respect to baseline period as shown in Fig. 11(d-f). The mean temperature rises above 3 °C over parts of western India and peninsular region.

Compared to the baseline period, the future mosquito density shows an increase by 20 to 60% in all the regions with maximum above 60% over the Western and peninsular regions by end of the century (Fig. 11 (g-i)). The increase is widespread and persists in near (2030s) , short term (2050s) and long term (2080s) future periods.

The percentage change in EIR shows an overall decrease by 10 to 30% in most of the regions in all the future projection periods compared to the baseline period but over western region and the peninsular region, the EIR values are projected to increase by 10 to 40% by end of the century (Fig. 11(j-l)).

Projections during OND

During north east monsoon season, rainfall is projected to increase over all the Indian regions. Especially, over the western and central region, rainfall is projected to increase above 40% by end of the century when compared to the baseline (Fig. 12 (a-c)). In other parts of India, there is a minimal rainfall increase by 0 to 20% is seen by the century end.

Fig. 12
figure 12

Same as Fig. 10 but for OND season.[NCAR Command Language (NCL version 6.6.2) is used to generate this figure].

Full size image

During OND, the future projections of mean temperature over most of the Indian regions show an increase of 0.5 to 1.5 °C, 1.5 to 2.5 °C and 2.5 to 4 °C for the periods 2030s, 2050s and 2080s, respectively, with respect to baseline period as shown in Fig. 12(d-f).The mean temperature is rising more in OND that JJAS season. Maximum temperature rise is seen over western and central India.

The mosquito density shows a maximum increase of 80% in western, central and western peninsular India by end of the century when compared to the baseline period while central India and eastern peninsular region show minimum increase of 30 to 60% for the same projection period (Fig. 12(g-i)).

The percentage change in EIR in all the projection period during OND shows an overall decrease in all the regions except over parts of western and peninsular region where the EIR shows an increase when compared to the baseline period. The EIR values are projected to increase above 40% by end of the century over these two regions where as in other regions, the EIR shows decrease of 10 to 30% (Fig. 12(j-l)).

Summary and conclusions

In this paper, the impact of climate change is estimated using VECTRI model and two common malaria indicators such as vector density and EIR are examined to predict the future malaria transmission over four homogeneous regions in India i.e western,central, eastren and peninsular India. Bias corrected daily temperature and rainfall data are from the multi model mean (MMM) of different CMIP6 global climate models with population data are used as inputs to VECTRI model and the simulation is done for the baseline period (1985–2014) as well as for three future projection periods of 21st century i.e. 2030s (2015-2035), 2050s (2035-2065), and 2080s (2065-2095). Results indicate that with the increase in temperature and rainfall, malaria transmission risk decreases over most of the India region during the twenty-first century (for near, mid and long term future projections) while it decreases over the peninsular India and parts of western India based on the recent IPCC global climate model projections under SSP5-8.5 scenario. Further, northwest monsoon season (OND) shows more intense in malaria transmission than southwest monsoon season (JJAS) over these regions. The percentage change in EIR in all the projection period during OND shows an overall increase by 40% over parts of western and peninsular region by end of the century when compared to the baseline period where as in other regions, the EIR shows decrease of 10 to 30% for the same period. Our results are found to be consistent with previous few studies which indicate a overall decrease in malaria transmission in Indian region, except over few regions like western and peninsular region, where the malaria transmission is increasing in future. As future rainfall shows higher increases in monsoon season over parts of central and eastern regions of India when compared to the historical period under SSP5-8.5 scenario. This increase in rainfall would flush off the existing breeding sites and would decrease early-stage larvae through flushing over these areas leading to reduction in EIR. Again, during post monsoon season (October to December), temperaure and rainfall makes the conditions favourable for the reproduction and growth of malaria vector and therefore EIR and vector density are higher during Oct–December months over peninsular region and western region with the highest value seen in the month of October. This is an attempt to enhance the understanding of future malaria transmission dynamics in India under SSP5-8.5 climate scenario, however, our results require further research using high resolution climate models under different green house gas emissions and malaria models with different mosquito diseases transmission mechanisms.

VECTRI default temperature thresholds for A. culicifacies species in India are set based on few previous studies who found similar thermal response curves for malaria vectors globally. While there is a requirement of comprehensive studies on parameterization of adult vector dynamics of malaria vector in India, specific to A. culicifacies, our modeling approach prioritized those parameters which are sensitive to climate change and climate variability in the present as well as in the future and its potential impacts on malaria transmission. Our model validation results demonstrate that this approach successfully captures the essential dynamics of malaria transmission in India. In future work, we intend to further refine the model by incorporating more species-specific parameters for A. culicifacies, particularly those related to adult vector dynamics and human-vector interactions. Also, our study does not consider changes in the malaria cases and the development changes in socio-economic factors that significantly impact malaria burden at regional scale. Our future study on malaria using VECTRI model may focus on this aspect.