Disease Forecasting Models Research Articles

Gradient boosting: A computationally efficient alternative to Markov chain Monte Carlo sampling for fitting large Bayesian spatio-temporal binomial regression models

Disease forecasting and surveillance often involve fitting models to a tremendous volume of historical testing data collected over space and time. Bayesian spatio-temporal regression models fit with Markov chain Monte Carlo (MCMC) methods are commonly used for such data. When the spatio-temporal support of the model is large, implementing an MCMC algorithm becomes a significant computational burden. This research proposes a computationally efficient gradient boosting algorithm for fitting a Bayesian spatio-temporal mixed effects binomial regression model. We demonstrate our method on a disease forecasting model and compare it to a computationally optimized MCMC approach. Both methods are used to produce monthly forecasts for Lyme disease, anaplasmosis, ehrlichiosis, and heartworm disease in domestic dogs for the contiguous United States. The data have a spatial support of 3108 counties and a temporal support of 108–138 months with 71–135 million test results. The proposed estimation approach is several orders of magnitude faster than the optimized MCMC algorithm, with a similar mean absolute prediction error.

Exploring Climate-Disease Connections in Geopolitical Versus Ecological Regions: The Case of West Nile Virus in the United States.

Many infectious disease forecasting models in the United States (US) are built with data partitioned into geopolitical regions centered on human activity as opposed to regions defined by natural ecosystems; although useful for data collection and intervention, this has the potential to mask biological relationships between the environment and disease. We explored this concept by analyzing the correlations between climate and West Nile virus (WNV) case data aggregated to geopolitical and ecological regions. We compared correlations between minimum, maximum, and mean annual temperature; precipitation; and annual WNV neuroinvasive disease (WNND) case data from 2005 to 2019 when partitioned into (a) climate regions defined by the National Oceanic and Atmospheric Administration (NOAA) and (b) Level I ecoregions defined by the Environmental Protection Agency (EPA). We found that correlations between climate and WNND in NOAA climate regions and EPA ecoregions were often contradictory in both direction and magnitude, with EPA ecoregions more often supporting previously established biological hypotheses and environmental dynamics underlying vector-borne disease transmission. Using ecological regions to examine the relationships between climate and disease cases can enhance the predictive power of forecasts at various scales, motivating a conceptual shift in large-scale analyses from geopolitical frameworks to more ecologically meaningful regions.

I-PomDiagnoser: A Real-Time Pomegranate Disease Management System

Objectives: Designing and developing i-PomDiagnoser: a real-time pomegranate disease management system for disease detection, classification, prediction, recommending preventive measures, and analyzing abrupt climatic changes and their impact on pomegranates. Methods: A data collection framework has been designed and developed using an agriculture drone, sensors, camera, and other equipment to collect real field pomegranate images and micro-level parameters. Comprehensive Exploratory Data Analysis (EDA) and Feature Selection (FS) processes were carried out to improve the accuracy of disease classification and forecasting models. ML-based Binary, Multimodel, and Multilabel classifiers were implemented for disease classification. The models were trained on 11 years of historical data and tested on 5 months of actual field data. A hybrid pomegranate disease forecasting model has been developed for accurately forecasting micro-level parameters for the next 45 days to predict diseases. Findings: Micro-level (weather, soil, water) parameters specific to the agro-climatic zone were collected. The five most prominent distinct diseases are considered for experimentation namely Bacterial Blight (Telya), Anthracnose, Fruit spot, Fusarium Wilt, and Fruit borer. The proposed Improved Ensemble Multilabel Classifier (i-Ensemble-MLC) with a modified voting scheme has achieved a high classification accuracy of 95.82%, addressing model overfitting and data imbalance. Moreover, the hybrid pomegranate disease forecasting model, combining LSTM and i-Ensemble-MLC, demonstrated better performance with minimal error rates (MSE: 0.003, RMSE: 0.071, MAE: 0.048, R2: 0.7) compared to the existing model1 (MSE:0.037, MAE:0.028). Novelty: The novelty lies in the creation of the all-in-one model, i-PomDiagnoser. This innovative system helps the farmers to correctly detect and predict the most prominent diseases of pomegranate. Keywords: Pomegranate, Agriculture, Disease Forecasting, Machine Learning, Deep Learning

Integrating Biological Control Agents for Enhanced Management of Apple Scab (Venturia inaequalis): Insights, Risks, Challenges, and Prospects

Apple scab incited by the ascomycete Venturia inaequalis poses a significant threat to apple cultivation, necessitating a reassessment of existing disease management strategies. Attempts to manage apple scab include diverse approaches like developing disease forecasting models and the extensive application of synthetic chemical fungicides. However, the efficacy of these methods is compromised by inconsistencies, environmental concerns, and the pathogen’s resistance, necessitating the exploration of alternative sustainable strategies. Addressing the challenges associated with apple scab management, this review strongly supports a shift towards the integration of biological control agents (BCAs). Emphasising the transformative synergy between BCAs and their bioactive secondary metabolites, we highlight their efficacy in advancing precision disease control through innovative and sustainable solutions. The review effectively presents a strong justification for the integration of BCAs and their by-products into apple scab management, offering insights into associated benefits, risks, and challenges while outlining promising prospects. Ultimately, it is expected to drive the adoption of environmentally conscious practices for effective apple scab management.

Scheduling Fungicide Applications for Cucurbit Downy Mildew Control on Pickling Cucumber in Michigan Using Disease Forecasters

Cucumber production is important to Michigan's economy and was valued at more than $45 million in 2019. Cucurbit downy mildew (CDM), caused by Pseudoperonospora cubensis, is an annual threat to Michigan's cucumber production, and fungicides must be applied frequently to prevent major yield losses. Our objective was to evaluate the disease forecasting models TOM-CAST, BLITE-CAST, and DM-CAST for their application in scheduling fungicide applications for CDM control. Field trials were conducted in 2021 and 2022 to evaluate each disease forecaster at different spray thresholds compared with 7- and 10-day programs and an untreated control. In 2021, all treatments received applications of cyazofamid alternated with oxathiapiprolin/chlorothalonil alternated with ametoctradin/dimethomorph plus chlorothalonil. The 2022 fungicide program was similar to that of 2021 except that cyazofamid was tank-mixed with chlorothalonil. Treatment plots were visually assessed for the foliar area (%) with CDM symptoms, and the relative area under the disease progress curve was determined at the end of each season. The results indicate that using DM-CAST or BLITE-CAST to schedule fungicide applications limited CDM and was similar to the 7-day program. The 7-day program received 7 (2021) and 6 (2022) applications, whereas DM-CAST and BLITE-CAST required 4 to 8 or 5 to 6 applications, respectively, depending on the threshold and year. This is the first study to evaluate these disease forecasters for scheduling fungicide application intervals for CDM.

Challenges and Opportunities in Plant Disease Management: A Brief Review

Plant diseases can reduce human food availability. Modern plant disease management faces problems due to climate change, fungicide resistance, pesticide residues and biodiversity loss. This review discusses problems and challenges in plant disease management and future research needs for effective management. Plant disease forecasting models can be used to predict plant diseases ahead of time. Protected cultivation combats climate change. It is necessary to find more evidence that plant pathogen diversity has a dilution effect on disease incidence. Deep learning-based disease diagnosis will help detect diseases faster. More hybrid fungicides should be developed to minimize fungicidal resistance problems. Among the molecular methods of plant disease management, RNA interference-mediated gene silencing and genome editing with CRISPR is more promising in plant disease management.

Epidemiology and control of strawberry powdery mildew: a review

Strawberry powdery mildew, caused by Podosphaera aphanis, is an economically important disease for strawberry production. Typical symptoms are white mycelium on all aerial parts of affected plants, with young host tissues being the most susceptible. The pathogen overwinters on infected leaves, either as mycelium or chasmothecia, although the quantitative role of chasmothecia in epidemics are not fully understood. In spring, under favourable conditions, the fungus sporulates, disseminating conidia and causing polycyclic infections. The disease is mainly controlled using synthetic fungicides, but there is increasing interest in sustainable alternatives, including microbial biocontrol agents (e.g., Ampelomyces quisqualis, Bacillus spp., Trichoderma spp.) and substances of plant or animal origin (e.g., Equisetum arvense, orange oil, chitosan, whey). Physical methods, (e.g. UV-C, ozone) are also promising alternatives to fungicides. All of these strategies should be combined with appropriate agronomic practices (e.g., overhead irrigation, canopy management) to create unfavourable environments for the pathogen. However, agronomic practices have never been assessed for P. aphanis. Disease forecasting models and DSSs, though available, are underutilized due to their complexity and lack of validation across locations. This review presents the current state of knowledge on P. aphanis the available methods for control of strawberry powdery mildew, and highlights knowledge gaps relating to this host/pathogen relationship.

Assessing the value of integrating national longitudinal shopping data into respiratory disease forecasting models

The COVID-19 pandemic led to unparalleled pressure on healthcare services. Improved healthcare planning in relation to diseases affecting the respiratory system has consequently become a key concern. We investigated the value of integrating sales of non-prescription medications commonly bought for managing respiratory symptoms, to improve forecasting of weekly registered deaths from respiratory disease at local levels across England, by using over 2 billion transactions logged by a UK high street retailer from March 2016 to March 2020. We report the results from the novel AI (Artificial Intelligence) explainability variable importance tool Model Class Reliance implemented on the PADRUS model (Prediction of Amount of Deaths by Respiratory disease Using Sales). PADRUS is a machine learning model optimised to predict registered deaths from respiratory disease in 314 local authority areas across England through the integration of shopping sales data and focused on purchases of non-prescription medications. We found strong evidence that models incorporating sales data significantly out-perform other models that solely use variables traditionally associated with respiratory disease (e.g. sociodemographics and weather data). Accuracy gains are highest (increases in R2 (coefficient of determination) between 0.09 to 0.11) in periods of maximum risk to the general public. Results demonstrate the potential to utilise sales data to monitor population health with information at a high level of geographic granularity.

Recent advances on pathogenesis, diagnosis, prevention, immunological aspects, and vectors of dengue: A review

Dengue is a vector-borne disease caused by the dengue virus (DENV) of family Flaviviridae. Dengue fever is common in both developed and developing countries. Globally, approximately 400 million cases of dengue fever are reported annually, resulting in approximately 22000 fatalities. Dengue cases in India have progressively increased in the last decade. In recent years, substantial progress has been made in understanding various aspects of dengue, including its pathogenesis, diagnosis, prevention strategies, immunological responses, and the role of vectors in its transmission. The transmission of a positive RNA virus occurs through Aedes mosquitoes, specifically Aedes aegypti and Aedes albopictus. This virus is associated with a wide spectrum of symptoms, ranging from mild undifferentiated fever to severe hemorrhagic fever and shock, posing a potential threat to human health. There are 4 types of antigenically distinct dengue serotypes (DENV-1 to DENV-4) and among them, DENV-2 is more lethal and extremely severe. To overcome the severity of dengue, Dengvaxia is administered to children 9 to 16 years old with evidence of previous dengue infection. The diagnosis of dengue is carried out by ELISA-based non-structural protein (NS1) and immunoglobulin tests. However, there are no specific biomarkers to identify severe disease progression. Climatic factors and temperature play an important role in complex interaction among host, vector, and virus to manifest the severity of dengue. There is a need for the refinement of climate-based disease forecasting models in India to effectively control the spread of dengue. The mosquito repellent should be used periodically to kill or repel the Aedes mosquito to prevent the spread of dengue in humans.

The Use of Digital Healthcare Systems to Predict Diseases

Smart health care depends heavily on a resilient and strong digital infrastructure. Telemedicine, Electronic Health Records (EHR), Fitness Trackers, Wearable Devices that monitor Heart Rate, Steps, Sleep Cycle and many other digital health-related measures are already used as indicators of what a future system of health technology will look like. The purpose of this paper is to examine the existing research studies to determine if it is possible to forecast health based on the data available from such devices. Further, in the Indian context, where Unique Health ID is already being implemented, this paper aims to extend the functionality of the UHID and analyze the viability of integrating the UHID with data sources for predicting health. Predicting and forecasting health will benefit all stakeholders in the healthcare ecosystem. Accurate disease forecasting models would be extremely helpful for epidemic and pandemic prevention and control. This research examines the potential for health forecasting and the challenges associated with its development.

Joint COVID-19 and influenza-like illness forecasts in the United States using internet search information

BackgroundAs the prolonged COVID-19 pandemic continues, severe seasonal Influenza (flu) may happen alongside COVID-19. This could cause a “twindemic”, in which there are additional burdens on health care resources and public safety compared to those occurring in the presence of a single infection. Amidst the raising trend of co-infections of the two diseases, forecasting both Influenza-like Illness (ILI) outbreaks and COVID-19 waves in a reliable and timely manner becomes more urgent than ever. Accurate and real-time joint prediction of the twindemic aids public health organizations and policymakers in adequate preparation and decision making. However, in the current pandemic, existing ILI and COVID-19 forecasting models face shortcomings under complex inter-disease dynamics, particularly due to the similarities in symptoms and healthcare-seeking patterns of the two diseases.MethodsInspired by the interconnection between ILI and COVID-19 activities, we combine related internet search and bi-disease time series information for the U.S. national level and state level forecasts. Our proposed ARGOX-Joint-Ensemble adopts a new ensemble framework that integrates ILI and COVID-19 disease forecasting models to pool the information between the two diseases and provide joint multi-resolution and multi-target predictions. Through a winner-takes-all ensemble fashion, our framework is able to adaptively select the most predictive COVID-19 or ILI signals.ResultsIn the retrospective evaluation, our model steadily outperforms alternative benchmark methods, and remains competitive with other publicly available models in both point estimates and probabilistic predictions (including intervals).ConclusionsThe success of our approach illustrates that pooling information between the ILI and COVID-19 leads to improved forecasting models than individual models for either of the disease.

Monitoring Spore Dispersal and Early Infections of Diplocarpon coronariae Causing Apple Blotch Using Spore Traps and a New qPCR Method.

Apple blotch (AB) is a major disease of apple in Asia and recently emerged in Europe and the United States. It is caused by the fungus Diplocarpon coronariae (formerly Marssonina coronaria; teleomorph: Diplocarpon mali) and leads to severe defoliation of apple trees in late summer, resulting in reduced yield and fruit quality. To develop effective disease management strategies, a sound knowledge of the pathogen's biology is crucial. Data on the early phase of disease development are scarce: No data on spore dispersal in Europe are available. We developed a highly sensitive TaqMan qPCR method to quantify D. coronariae conidia in spore trap samples. We monitored temporal and spatial dispersal of conidia of D. coronariae and the progress of AB in spring and early summer in an extensively managed apple orchard in Switzerland in 2019 and 2020. Our results show that D. coronariae overwinters in leaf litter, and spore dispersal and primary infections occur in late April and early May. We provide the first results describing early-season dispersal ofconidia of D. coronariae, which, combined with the observed disease progress, helps to understand the disease dynamics and will be a basis for improved disease forecast models. Using the new qPCR method, we detected D. coronariae in buds, on bark, and on fruit mummies, suggesting that several apple tissues might serve as overwintering habitats for the fungus, in addition to fallen leaves. [Formula: see text] Copyright © 2023 The Author(s). This is an open access article distributed under the CC BY-NC-ND 4.0 International license.

Disease Forecasting for the Rational Management of Grapevine Mildews in the Chianti Bio-District (Tuscany)

Downy and powdery mildews are major grapevine diseases. In organic viticulture, a few fungicides with protectant activities (copper and sulphur in particular) can be used, and their preventative application frequently leads to unneeded spraying. The adoption of an epidemiological disease forecasting model could optimise the timing of treatments and achieve a good level of disease protection. In this study, the effectiveness of the EPI (Etat Potentiel d'Infection) model in predicting infection risk for downy and powdery mildews was evaluated in nine organic vineyards located in Panzano in Chianti (FI), over a 2-year period (2020-2021). The reliability of the EPI model was investigated by comparing the disease intensities, the number of fungicide sprayings, the quantities of the fungicides (kg/ha), and the costs of the treatment achieved, with or without the use of the model, in a vineyard. The results obtained over two seasons indicated that, in most cases, the use of the EPI model accurately signalled the infection risk and allowed for a reduction in the frequency and cost of spraying, particularly for powdery mildew control (-40% sprayings, -20% costs compared to the farmer's schedule), without compromising crop protection. The use of the EPI model can, therefore, contribute to more-sustainable disease management in organic viticulture.

The effect of temperature, leaf wetness period, and cultivar susceptibility on boxwood blight disease development and sporulation.

Boxwood blight causes great losses to the boxwood nursery industry and landscapes in 30 states in the USA. Understanding the epidemiological factors governing disease development will be important for disease forecasting and design of best management practices. We evaluated the effect of leaf wetness period (lwp) and temperature on lesion development and sporulation on three boxwood cultivars under controlled conditions to develop predictive models for disease development. We conducted detached leaf assays at 18 to 27°C, and various lwp, with the cultivars Buxus sempervirens 'Suffruticosa' (highly susceptible), Buxus × 'Green Velvet' (moderately susceptible), and B. microphylla var. japonica 'Winter Gem' (less susceptible). Detached leaves were inoculated with 200 conidia in 50 µL of suspension and disease incidence was recorded at 3 - 13 d post inoculation (dpi). Cultivar, lwp, temperature, and most interactions significantly influenced disease development. A minimum of 5 h of leaf wetness was required for any disease. Lesion development increased most rapidly between 12 - 15 h and continued to increase to about 21 h of leaf wetness. Temperatures between 21 and 25°C were optimal for lesion development. There was about a 7-d lag between appearance of lesions and maximal incidence of sporulation. The two less-susceptible cultivars had fewer lesions than 'Suffruticosa' under the same infection conditions; in addition, leaf lesions of 'Winter Gem' exhibited delayed sporulation and sporulation from a smaller proportion of symptomatic leaves. Response surfaces were developed for each cultivar to predict the disease incidence using the lwp and dpi. Our findings will help refine disease forecast models to improve management of boxwood blight.

Development of Botrytis cinerea under reduction of pesticides treatments in Macedonian viticulture production

The monitoring of development of Botrytis cinerea under reduction of pesticide treatments in Macedonian viticulture production is possible only when introducing a disease forecasting model. B.cinerea causes an increase in the number of chemical treatments just before harvest and calls into question the environmental and health value of the product. Therefore, an attempt was made to create a forecasting model for Botrytis which is based on the relationship between relative humidity and the temperature in the vine canopy. The forecasting model for Botrytis was applied at the white varieties Smederevka and Zilavka and based on the data obtained was made ANOVA statistical test which proves the reliability of the model. On the localities, Smilica and Sopot, Kavadarci, Republic of Macedonia, are the experimental fields that were observed for three consecutive years (2017 till 2019).

Stepwise Regression Models-Based Prediction for Leaf Rust Severity and Yield Loss in Wheat

Leaf rust is a devastating disease in wheat crop. The disease forecasting models can facilitate the economic and effective use of fungicides and assist in limiting crop yield losses. In this study, six wheat cultivars were screened against leaf rust at two locations, during three consecutive growing seasons. Subsequently, the stepwise regression analysis was employed to analyze the correlation of six epidemiological variables (minimum temperature, maximum temperature, minimum relative humidity, maximum relative humidity, rainfall and wind speed) with disease severity and yield loss (%). Disease predictive models were developed for each cultivar for final leaf rust severity and yield loss prediction. Principally, all epidemiological variables indicated a positive association with leaf rust severity and yield loss (%) except minimum relative humidity. The effectiveness of disease predictive models was estimated using coefficient of determination (R2) values for all models. Then, these predictive models were validated to forecast disease severity and yield loss at another location in Faisalabad. The R2 values of all disease predictive models for each of the tested cultivars were high, evincing that our regression models could be effectively employed to predict leaf rust disease severity and anticipated yield loss. The validation results explained 99% variability, suggesting a highly accurate prediction of the two variables (leaf rust severity and yield loss). The models developed in this research can be used by wheat farmers to forecast disease epidemics and to make disease management decisions accordingly.

Isolation and Identification of the Causal Agents of Red Pepper Wilting Symptoms

In order to investigate the cause of wilting symptoms in red pepper field of Korea, the frequency of occurrence of red peppers showing wilting symptoms was investigated in pepper cultivation fields in Goesan, Chungcheongbuk-do for 5 years from 2010 to 2014. There was a difference in the frequency of wilting symptoms depending on the year of investigation, but the frequency of occurrence increased as the investigation period passed from June and July to August. During this period, Ralstonia solanacearum causing the bacterial wilt was isolated at a rate four times higher than Phytophthora capsica causing the Phytophthora late blight. In wilted peppers collected in Goesan of Chungbuk and Andong of Gyeongbuk in 2013 and 2014, R. solanacearum and P. capsici were isolated from 20.3% and 3.8% of the total fields, respectively. In the year with a high rate of wilting symptoms, the average temperature was high, and the disease occurrence date of the bacterial wilt, estimated with disease forecasting model, was also fast. The inconsistency between the number of days at risk of Phytophthora late blight and the frequency of occurrence of wither symptoms is thought to be due to the generalization of the use of cultivars resistant to the Phytophthora late blight in the pepper field. In our study, the wilting symptoms were caused by the bacterial wilt caused by R. solanacearum rather than the Phytophthora late blight caused by P. capsica, which is possibly caused by increasing cultivation of pepper varieties resistant to the Phytophthora late blight in the field.

A public decision support system for the assessment of plant disease infection risk shared by Italian regions.

Integrated pest management (IPM) practices proved to be efficient in reducing pesticide use and ensuring economic farming sustainability. Digital decision support systems (DSS) to support the adoption of IPM practices from plant protection services are required by European legislation. Available DSSs used by Italian plant protection services are heterogeneous with regards to disease forecasting models, datasets for their calibration, and level of integration in operational decision-making. This study presents the MISFITS-DSS, which has been jointly developed by a public research institution and nine regional plant protection services with the objective of harmonizing data collection and decision support for Italian farmers. Participatory approach allowed designing a predictive workflow relying on specific domain expertise, in order to explicitly match actual user needs. The DSS calibration entailed the risk of grapevine downy mildew infection (5-point scale from very low to very high), and phenological observations in 2012-2017 as reference data. Process-based models of primary and secondary infections have been implemented and tested via sensitivity analysis (Morris method) under contrasting weather conditions. Hindcast simulations of grapevine phenology, host susceptibility and disease pressure were post-processed by machine-learning classifiers to predict the reference infection risk. Results indicate that IPM principles are implemented by plant protection services since years. The accurate reproduction of grapevine phenology (RMSE=4-14 days), which drove the dynamic of host susceptibility, and the use of weather forecasts as model inputs contributed to reliably predict the reference infection risk (88% balanced accuracy). We did a pioneering effort to homogenize the methodology to deliver decision support to Italian farmers, by involving plant protection services in the DSS definition, to foster a further adoption of IPM practices.

Prediction of mungbean yellow mosaic virus disease using multiple regression models

BackgroundThe main cause for low mungbean [Vigna radiata (L.) R. Wilczek] productivity is mungbean yellow mosaic virus (MYMV). Generally, the management of MYMV relies upon frequent insecticide sprays to control its vector (whitefly) and genetic resistance. However, disease forecast models can help to economize the pesticide sprays. Hence current study was designed to identifying environmental factor that promote disease development and developing a disease prediction model. MethodsOne hundred and twenty-seven mungbean accessions were planted for two years (2012 and 2013) and infection was dependent on natural inoculum. Weekly and daily data on disease incidence and environmental variables were collected and analyzed using correlation and stepwise regression analysis. ResultsWind velocity and high temperature had a negative relation with disease occurrence during both years, whereas low temperature, rainfall, and relative humidity, had positive relationship based on linear regression. The environmental conditions responsible for the highest disease incidence were, maximum temperature (32–34 °C), relative humidity (72–75 %), minimum temperature (27–29 °C), rainfall (1.8–2.1 mm) and wind velocity (3–4.5 km/hr) during both growing seasons Overall, five environmental variable multiple regression model encompassing relative humidity, wind speed, rainfall, suboptimum temperature, and optimum temperatures accommodate the data rightly explaining 83 % variation in disease outgrowth. ConclusionThe observed MYMV disease occurrence values for most of the mungbean genotypes, and those predicted by the model were very close. This multi-environmental variable model can be utilized to provide early warning forecasts for the management of MYMV in Pakistan.

Forecasting Alternaria Leaf Spot in Apple with Spatial-Temporal Meteorological and Mobile Internet-Based Disease Survey Data

Early warning of plant diseases and pests is critical to ensuring food safety and production for economic crops. Data sources such as the occurrence, frequency, and infection locations are crucial in forecasting plant diseases and pests. However, at present, acquiring such data relies on fixed-point observations or field experiments run by agricultural institutions. Thus, insufficient data and low rates of regional representative are among the major problems affecting the performance of forecasting models. In recent years, the development of mobile internet technology and conveniently accessible multi-source agricultural information bring new ideas to plant diseases’ and pests’ forecasting. This study proposed a forecasting model of Alternaria Leaf Spot (ALS) disease in apple that is based on mobile internet disease survey data and high resolution spatial-temporal meteorological data. Firstly, a mobile internet-based questionnaire was designed to collect disease survey data efficiently. A specific data clean procedure was proposed to mitigate the noise in the data. Next, a sensitivity analysis was performed on the temperature and humidity data, to identify disease-sensitive meteorological factors as model inputs. Finally, the disease forecasting model of the apple ALS was established using four machine learning algorithms: Logistic regression(LR); Fisher linear discriminant analysis(FLDA); Support vector machine(SVM); and K-Nearest Neighbors (KNN). The KNN algorithm is recommended in this study, which produced an overall accuracy of 88%, and Kappa of 0.53. This paper shows that through mobile internet disease survey and a proper data clean approach, it is possible to collect necessary data for disease forecasting in a short time. With the aid of high resolution spatial-temporal meteorological data and machine learning approaches, it is able to achieve disease forecast at a regional scale, which will facilitate efficient disease prevention practices.