Understanding the spatial non-stationarity in the relationships between malaria incidence and environmental risk factors using Geographically Weighted Random Forest: A case study in Rwanda.

As found in the health studies literature, the levels of climate association between epidemiological diseases have been found to vary across regions. Therefore, it seems reasonable to allow for the possibility that relationships might vary spatially within regions. We implemented the geographically weighted random forest (GWRF) machine learning method to analyze ecological disease patterns caused by spatially non-stationary processes using a malaria incidence dataset for Rwanda. We first compared the geographically weighted regression (WGR), the global random forest (GRF), and the geographically weighted random forest (GWRF) to examine the spatial non-stationarity in the non-linear relationships between malaria incidence and their risk factors. We used the Gaussian areal kriging model to disaggregate the malaria incidence at the local administrative cell level to understand the relationships at a fine scale since the model goodness of fit was not satisfactory to explain malaria incidence due to the limited number of sample values. Our results show that in terms of the coefficients of determination and prediction accuracy, the geographical random forest model performs better than the GWR and the global random forest model. The coefficients of determination of the geographically weighted regression (R2), the global RF (R2), and the GWRF (R2) were 4.74, 0.76, and 0.79, respectively. The GWRF algorithm achieves the best result and reveals that risk factors (rainfall, land surface temperature, elevation, and air temperature) have a strong non-linear relationship with the spatial distribution of malaria incidence rates, which could have implications for supporting local initiatives for malaria elimination in Rwanda.

Spatial Analysis of Ambient Air Pollution and Cardiovascular Disease (CVD) Hospitalization Across Sweden.

The associations of multiple pollutants and cardiovascular disease (CVD) morbidity, and the spatial variations of these associations have not been nationally studied in Sweden. The main aim of this study was, thus, to spatially analyze the associations between ambient air pollution (black carbon, carbon monoxide, particulate matter (both <10 µm and <2.5 µm in diameter) and Sulfur oxides considered) and CVD admissions while controlling for neighborhood deprivation across Sweden from 2005 to 2010. Annual emission estimates across Sweden along with admission records for coronary heart disease, ischemic stroke, atherosclerotic and aortic disease were obtained and aggregated at Small Areas for Market Statistics level. Global associations were analyzed using global Poisson regression and spatially autoregressive Poisson regression models. Spatial non-stationarity of the associations was analyzed using Geographically Weighted Poisson Regression. Generally, weak but significant associations were observed between most of the air pollutants and CVD admissions. These associations were non-homogeneous, with more variability in the southern parts of Sweden. Our study demonstrates significant spatially varying associations between ambient air pollution and CVD admissions across Sweden and provides an empirical basis for developing healthcare policies and intervention strategies with more emphasis on local impacts of ambient air pollution on CVD outcomes in Sweden.

Data for assessment of leached dissolved organic carbon in watersheds.

("Dissolved organic carbon leaching flux in a mixed agriculture and forest watershed in Rwanda" [1]). This article presents data of leached dissolved organic carbon (LDOC), stream water dissolved organic carbon), rainfall amount (Ra), rainfall intensity (Ri), rainfall soil storage (S), runoff (Q), and soil properties such as total organic carbon (TOC), total nitrogen (TN), cation exchange capacity (CEC), and soil texture data collected in the Rukarara River Watershed (RRW), a tropical watershed. All these data were used to analyze leached dissolved organic carbon (LDOC) fluxes in the watershed and their relationship with stream DOC. LDOC and soil properties data were collected at three sites in multiple plots per site located in natural forest (NF), tea plantations (TP), plantation forests (PF), and croplands (CL). Twenty-three plots in total were sampled to collect LDOC data. Soil properties data were analyzed from soil samples collected nearby the plots. Soil texture elements data were used to calculate soil porosity and saturated hydraulic conductivity (Ks). Data of stream DOC were analyzed from water samples collected and analyzed in the laboratory using a TOC analyzer. Rainfall data were recorded within the RRW using tipping bucket rain gauges installed at three sites. These rainfall data were used to calculate rainfall intensity, potential surface runoff, and rainfall soil storage.

Data for assessment of soil water extractable and percolation water dissolved organic carbon in watersheds.

("Sources of soil dissolved organic carbon in a mixed agricultural and forested watershed in Rwanda", [6]) This data article presents water extractable organic carbon (WEOC), percolation water dissolved organic carbon (pDOC), and mean antecedent precipitation indices (API) and mean antecedent temperature (MAT) data. The article also presents edaphic properties such soil texture elements, total organic carbon (TOC), total nitrogen (TN), cation exchange capacity (CEC), iron (Fe), and aluminum (Al). Additionally, the article presents topography attributes such including topographic position index (TPI) and curvature. All these data were used to analyze both WEOC and pDOC dynamics in the Rukarara River Watershed (RRW), Rwanda. WEOC and soil properties data were analyzed from sampled 52 soil composites samples collected during from October to December 2016 using 53 × 50 mm rings. Data of pDOC were analyzed from percolation water samples collected using a zero tension lysimeters on various dates during the period from Jun 2015 to Jun 2017. API and MAT data for various antecedent days were calculated on basis of rainfall and air temperature data recorded at three stations within the RRW using respectively tipping bucket rain gauges and sensors installed at three sites located representing the main land use land cover classes within the RRW.

Establishing spatially-enabled health registry systems using implicit spatial data pools: case study - Uganda.

BACKGROUND: Spatial epidemiological analyses primarily depend on spatially-indexed medical records. Some countries have devised ways of capturing patient-specific spatial details using ZIP codes, postcodes or personal numbers, which are geocoded. However, for most resource-constrained African countries, the absence of a means to capture patient resident location as well as inexistence of spatial data infrastructures makes capturing of patient-level spatial data unattainable. METHODS: This paper proposes and demonstrates a creative low-cost solution to address the issue. The solution is based on using interoperable web services to capture fine-scale locational information from existing "spatial data pools" and link them to the patients' information. RESULTS: Based on a case study in Uganda, the paper presents the idea and develops a prototype for a spatially-enabled health registry system that allows for fine-level spatial epidemiological analyses. CONCLUSION: It has been shown and discussed that the proposed solution is feasible for implementation and the collected spatially-indexed data can be used in spatial epidemiological analyses to identify hotspot areas with elevated disease incidence rates, link health outcomes to environmental exposures, and generally improve healthcare planning and provisioning.

Comment on "The global tree restoration potential".

Bastin et al (Reports, 5 July 2019, p. 76) claim that 205 gigatonnes of carbon can be globally sequestered by restoring 0.9 billion hectares of forest and woodland canopy cover. Reinterpreting the data from Bastin et al, we show that the global land area actually required to sequester human-emitted CO2 is at least a factor of 3 higher, representing an unrealistically large area.

Spatial analysis of HIV-TB co-clustering in Uganda.

BACKGROUND: Tuberculosis (TB) is the leading cause of death for individuals infected with Human immunodeficiency virus (HIV). Conversely, HIV is the most important risk factor in the progression of TB from the latent to the active status. In order to manage this double epidemic situation, an integrated approach that includes HIV management in TB patients was proposed by the World Health Organization and was implemented in Uganda (one of the countries endemic with both diseases). To enable targeted intervention using the integrated approach, areas with high disease prevalence rates for TB and HIV need to be identified first. However, there is no such study in Uganda, addressing the joint spatial patterns of these two diseases. METHODS: This study uses global Moran's index, spatial scan statistics and bivariate global and local Moran's indices to investigate the geographical clustering patterns of both diseases, as individuals and as combined. The data used are TB and HIV case data for 2015, 2016 and 2017 obtained from the District Health Information Software 2 system, housed and maintained by the Ministry of Health, Uganda. RESULTS: Results from this analysis show that while TB and HIV diseases are highly correlated (55-76%), they exhibit relatively different spatial clustering patterns across Uganda. The joint TB/HIV prevalence shows consistent hotspot clusters around districts surrounding Lake Victoria as well as northern Uganda. These two clusters could be linked to the presence of high HIV prevalence among the fishing communities of Lake Victoria and the presence of refugees and internally displaced people camps, respectively. The consistent cold spot observed in eastern Uganda and around Kasese could be explained by low HIV prevalence in communities with circumcision tradition. CONCLUSIONS: This study makes a significant contribution to TB/HIV public health bodies around Uganda by identifying areas with high joint disease burden, in the light of TB/HIV co-infection. It, thus, provides a valuable starting point for an informed and targeted intervention, as a positive step towards a TB and HIV-AIDS free community.

Predictive risk mapping of human leptospirosis using support vector machine classification and multilayer perceptron neural network.

Leptospirosis is a zoonotic disease found wherever human is in direct or indirect contact with contaminated water and environment. Considering the increasing number of cases of this disease in the northern part of Iran, identifying areas characterized by high disease incidence risk can help policy-makers develop strategies to prevent its further spread. This study presents an approach for generating predictive risk maps of leptospirosis using spatial statistics, environmental variables and machine learning. Moran's I demonstrated that the distribution of leptospirosis cases in the study area in Iran was highly clustered. Pearson's correlation analysis was conducted to examine the type and strength of relationships between climate and topographical factors and incidence of the disease. To handle the complex and nonlinear problems involved, machine learning based on the support vector machine classification algorithm and multilayer perceptron neural network was exploited to generate annual and monthly predictive risk maps of leptospirosis distribution. Performance of both models was evaluated using receiver operating characteristic curve and Kappa coefficient. The output results demonstrated that both models are adequate for the prediction of the probability of leptospirosis incidence.

Proposing and investigating PCAMARS as a novel model for NO2 interpolation.

Effective measurement of exposure to air pollution, not least NO2, for epidemiological studies along with the need to better management and control of air pollution in urban areas ask for precise interpolation and determination of the concentration of pollutants in nonmonitored spots. A variety of approaches have been developed and used. This paper aims to propose, develop, and test a spatial predictive model based on multivariate adaptive regression splines (MARS) and principle component analysis (PCA) to determine the concentration of NO2 in Tehran, as a case study. To increase the accuracy of the model, spatial data (population, road network and point of interests such as petroleum stations and green spaces) and meteorological data (including temperature, pressure, wind speed and relative humidity) have also been used as independent variables, alongside air quality measurement data gathered by the monitoring stations. The outputs of the proposed model are evaluated against reference interpolation techniques including inverse distance weighting, thin plate splines, kriging, cokriging, and MARS3. Interpolation for 12 months showed better accuracies of the proposed model in comparison with the reference methods.

Data for dynamics analysis of riverine dissolved organic in watersheds.

This data article presents water stage, flow, and net primary productivity (NPP) data that were used to analyze the dynamics of the riverine dissolved organic carbon (DOC) dynamics in the Rukarara River watershed in Rwanda. We measured water stage data every 15 min and calculated daily averages used to estimate flow based on rating curves. The rating curves were produced using several measured contemporaneous water stage and flow data. Estimated flow data were used to calculate water dissolved organic carbon (DOC) loadings separate the total stream flow into quick and baseflow. Annual NPP data for a 15-year period were used to estimate the effect of proportion of stream DOC loading on carbon sequestration within the Rukarara River watershed.

Dissolved organic carbon in streams within a subarctic catchment analysed using a GIS/remote sensing approach.

Climate change projections show that temperature and precipitation increases can alter the exchange of greenhouse gases between the atmosphere and high latitude landscapes, including their freshwaters. Dissolved organic carbon (DOC) plays an important role in greenhouse gas emissions, but the impact of catchment productivity on DOC release to subarctic waters remains poorly known, especially at regional scales. We test the hypothesis that increased terrestrial productivity, as indicated by the normalized difference vegetation index (NDVI), generates higher stream DOC concentrations in the Stordalen catchment in subarctic Sweden. Furthermore, we aimed to determine the degree to which other generic catchment properties (elevation, slope) explain DOC concentration, and whether or not land cover variables representing the local vegetation type (e.g., mire, forest) need to be included to obtain adequate predictive models for DOC delivered into rivers. We show that the land cover type, especially the proportion of mire, played a dominant role in the catchment's release of DOC, while NDVI, slope, and elevation were supporting predictor variables. The NDVI as a single predictor showed weak and inconsistent relationships to DOC concentrations in recipient waters, yet NDVI was a significant positive regulator of DOC in multiple regression models that included land cover variables. Our study illustrates that vegetation type exerts primary control in DOC regulation in Stordalen, while productivity (NDVI) is of secondary importance. Thus, predictive multiple linear regression models for DOC can be utilized combining these different types of explanatory variables.

Riverine dissolved organic carbon in Rukarara River Watershed, Rwanda.

Dissolved organic carbon (DOC) loading is rarely estimated in tropical watersheds. This study quantifies DOC loading in the Rukarara River Watershed (RRW), a Rwandan tropical forest and agricultural watershed, and evaluates its relationship with hydrological factors, land use and land cover (LULC), and topography to better understand the impact of stream DOC export on watershed carbon budgets. The annual average load for the study period was 977.80 kg C, which represents approximately 8.44% of the net primary productivity of the watershed. The mean daily exports were 0.37, 0.14, 0.075 and 0.32 kg C/m2 in streams located in natural forest, tea plantation, small farming areas, and at the outlet of the river, respectively. LULC is a factor that influences DOC loading. The quick flow was the main source of stream DOC at all study sites. Stream DOC increases with increasing water flow, indicating a positive relationship. Thus, the expectation is that a change in land cover and/or rainfall will result in a change of stream DOC dynamics within the watershed. Topography was also found to influence the dynamics of stream DOC through its effect on overland flow in terms of drainage area and total length of flow paths. Tea plantations were located in areas of high drainage density and projected increase of rainfall in the region, as a consequence of climate change, could increase stream DOC content and affect stream water quality, biodiversity, balance between autotrophy and heterotrophy, and bioavailability of toxic compounds within the RRW.

Exploring spatial patterns of cardiovascular disease in Sweden between 2000 and 2010.

AIMS: Cardiovascular disease (CVD) is one of the leading causes of mortality and morbidity worldwide, including in Sweden. The main aim of this study was to explore the temporal trends and spatial patterns of CVD in Sweden using spatial autocorrelation analyses. METHODS: The CVD admission rates between 2000 and 2010 throughout Sweden were entered as the input disease data for the analytic processes performed for the Swedish capital, Stockholm, and also for the whole of Sweden. Age-adjusted admission rates were calculated using a direct standardisation approach for men and women, and temporal trends analysis were performed on the standardised rates. Global Moran's I was used to explore the structure of patterns and Anselin's local Moran's I, together with Kulldorff's scan statistic were applied to explore the geographical patterns of admission rates. RESULTS: The rates followed a spatially clustered pattern in Sweden with differences occurring between sexes. Accordingly, hot spots were identified in northern Sweden, with higher intensity identified for men, together with clusters in central Sweden. Cold spots were identified in the adjacency of the three major Swedish cities of Stockholm, Gothenburg and Malmö. CONCLUSIONS: The findings of this study can serve as a basis for distribution of health-care resources, preventive measures and exploration of aetiological factors.

Drivers of dissolved organic carbon export in a subarctic catchment: Importance of microbial decomposition, sorption-desorption, peatland and lateral flow.

Tundra soils account for 50% of global stocks of soil organic carbon (SOC), and it is expected that the amplified climate warming in high latitude could cause loss of this SOC through decomposition. Decomposed SOC could become hydrologically accessible, which increase downstream dissolved organic carbon (DOC) export and subsequent carbon release to the atmosphere, constituting a positive feedback to climate warming. However, DOC export is often neglected in ecosystem models. In this paper, we incorporate processes related to DOC production, mineralization, diffusion, sorption-desorption, and leaching into a customized arctic version of the dynamic ecosystem model LPJ-GUESS in order to mechanistically model catchment DOC export, and to link this flux to other ecosystem processes. The extended LPJ-GUESS is compared to observed DOC export at Stordalen catchment in northern Sweden. Vegetation communities include flood-tolerant graminoids (Eriophorum) and Sphagnum moss, birch forest and dwarf shrub communities. The processes, sorption-desorption and microbial decomposition (DOC production and mineralization) are found to contribute most to the variance in DOC export based on a detailed variance-based Sobol sensitivity analysis (SA) at grid cell-level. Catchment-level SA shows that the highest mean DOC exports come from the Eriophorum peatland (fen). A comparison with observations shows that the model captures the seasonality of DOC fluxes. Two catchment simulations, one without water lateral routing and one without peatland processes, were compared with the catchment simulations with all processes. The comparison showed that the current implementation of catchment lateral flow and peatland processes in LPJ-GUESS are essential to capture catchment-level DOC dynamics and indicate the model is at an appropriate level of complexity to represent the main mechanism of DOC dynamics in soils. The extended model provides a new tool to investigate potential interactions among climate change, vegetation dynamics, soil hydrology and DOC dynamics at both stand-alone to catchment scales.

Cropland changes in times of conflict, reconstruction, and economic development in Iraqi Kurdistan.

The destruction of land and forced migration during the Anfal attacks against the Kurds in Iraq in the late 1980s has been reported to have severe consequences for agricultural development. A reconstruction program to aid people in returning to their lands was launched in 1991. To assess the agricultural situation in the Duhok governorate during the pre-Anfal (A), post-Anfal (B), reconstruction (C), and present (D) periods, we mapped winter crops by focusing on inter-annual variability in vegetation greenness, using satellite images. The results indicate a decrease in cultivated area between period A and B, and a small increase between period B and C. This supports reports of a decline in cultivated area related to the Anfal campaign, and indicates increased activity during the reconstruction program. Period D showed a potential recovery with a cropland area similar to period A.

Environmental modelling of visceral leishmaniasis by susceptibility-mapping using neural networks: a case study in north-western Iran.

Visceral leishmaniasis (VL) is a potentially fatal vector-borne zoonotic disease, which has become an increasing public health problem in the north-western part of Iran. This work presents an environmental health modelling approach to map the potential of VL outbreaks in this part of the country. Radial basis functional link networks is used as a data-driven method for predictive mapping of VL in the study area. The high susceptibility areas for VL outbreaks account for 36.3% of the study area and occur mainly in the north (which may affect the neighbouring countries) and South (which is a warning for other provinces in Iran). These parts of the study area have many nomadic, riverside villages. The overall accuracy of the resultant map was 92% in endemic villages. Such susceptibility maps can be used as reconnaissance guides for planning of effective control strategies and identification of possible new VL endemic areas.

Are associations between socio-economic characteristics and exposure to air pollution a question of study area size? An example from Scania, Sweden.

BACKGROUND: Numerous studies have shown that exposure to air pollutants in the area of residence and the socio-economic status of an individual may be related. Therefore, when conducting an epidemiological study on the health effect of air pollution, socio-economy may act as a confounding factor. In this paper we examine to what extent socio-economic status and concentrations of NO2 in the county/region of Scania, southern Sweden, are associated and if such associations between these factors differ when studying them at county or city level. To perform this study we used high-resolution census data and modelled the annual exposure to NO2 using an emission database, a dispersion modelling program and a geographical information system (GIS). RESULTS: The results from this study confirm that socio-economic status and the levels of NO2 in the area of residence are associated in some cities. The associations vary considerably between cities within the same county (Scania). Even for cities of similar sizes and population bases the associations observed are different. Studying the cities together or separately yields contradictory results, especially when education is used as a socio-economic indicator. CONCLUSION: Four conclusions have been drawn from the results of this study. 1) Adjusting for socio-economy is important when investigating the health effects of air pollution. 2) The county of Scania seems to be heterogeneous regarding the association between air pollution and socio-economy. 3) The relationship between air pollution and socio-economy differs in the five cities included in our study, depending on whether they are analysed separately or together. It is therefore inadvisable to determine and analyse associations between socio-economy and exposure to air pollutants on county level. This study indicates that the size and choice of study area is of great importance. 4) The selection of socio-economic indices (in this study: country of birth and education level) is important.