So now, let's turn to malaria and climate change. So increasing temperature could result in faster development of the mosquito from egg to adult, decreased interval between mosquito's blood meals, faster development of the parasite in mosquito, longer duration of the transmission season, spread to higher altitudes and latitudes. Leave a question mark there but I'm going to show some work regarding that. Increasing rainfall could lead to pools of water to serve as larval development sites. Finally, hot and dry conditions could result in decreased mosquito survival. Okay. So now let's look at a very well done study entitled Altitudinal Changes in Malaria Incidence in Highlands of Ethiopia and Colombia. So this was a high-quality observational study. The research question was, "How does temperature variability influence the spatial distribution of malaria incidence along altitudinal gradients?" This was a case where there was high-quality long-terms time-series data available, in this case, for monthly microscopically-confirmed Plasmodium falciparum cases. In Colombia, the data were available for 124 municipalities in a particular region from 1990-2006. In Ethiopia, the data were available for 159 kebeles, which is their name for municipalities, in a particular region for 1993-2004. They had altitude data that was satellite-based, based on the Advanced Spaceborne Thermal Emission and Reflection Radiometer Global Digital Elevation Model, which is quite a mouthful. They had temperature data from four meteorological stations in Ethiopia and nine in Colombia and, of course, those were situated in close proximity to the study areas. So these are the geographic locations for the two study areas. Colombia, Ethiopia. Here, this is the region studied in Ethiopia; here, we see the regions studied in Colombia. This shows the altitude scale, and you could see that both regions encompassed a range of altitude, which was necessary to have a successful study. Okay. Now, let's look at, first, relatively simple result. I'll walk you through this. Ethiopia, Colombia. In each of these graphs, the x-axis is altitude, and the y-axis is the log of the number of cases per 10,000 population. Here, we're looking at 1994-1997, and then all years. To make a long story short, you could see that in all cases, we see a decreasing incidence with increasing altitude. So that's very simple descriptive data. Okay. Then they did something more sophisticated, where they used a statistical method to group the municipalities according to their temporal dynamics of malaria cases. So they created three groups. Group 1, 2, and 3, with Group 1 the red. Sorry, let me [inaudible] more. Ethiopia, Colombia. Here, we're looking at the group distribution. So the Group 1, the red, had the highest malaria incidence, and the yellow had the lowest malaria incidence. So here on this side, we're showing the altitude gradient, with yellow being high altitude, the brown being lower altitude. So what you see tracks very nicely is that we're seeing the high incidence correspond very well with the low altitudes. Here also. Maybe not quite as well here as here. We see the low incidence, the yellow, correspond quite well with the high altitudes, where, of course, I should have mentioned before, but maybe it was implicit that the higher the altitude, the cooler the temperature; the lower the altitude, the warmer the temperature. Okay. So this analysis is more complicated. I'll walk you through it. Again, Ethiopia, Colombia. They defined median altitude as the altitude at which 50 percent of the cases occur below and 50 percent occurred above that altitude. So here, we're looking at altitude and the cumulative number of cases. In 1994, the altitude at which 50 percent of the cases occurred was here, which is a little over 1,800 meters. So that's considered the median altitude in 1994. In 1997, you can see that it's shifted. So here's where we have 50 percent of the cases occurring. Now the median altitude maybe it's hard to tell, let's say 1,880 meters or so. You could see a more dramatic difference between 1994 and 1997 in Colombia. So 1994, here's where you see 50 percent of the cases occurring, and that current at an altitude of around 1,300 meters, then in 1997, it was almost 1,600 meters. Okay. So that shows you how they calculated the median altitude. Okay. So then over here, where we're looking at is the x-axis is mean temperature, and the y-axis is median altitude. So the mean temperature is the mean temperature for the entire region, either Ethiopia or Columbia. It's for the four months preceding the onset of the malaria season because other work determined that it's actually the four months preceding the season that's important in determining the malaria incidence for that particular season. So what we see here is that with increasing temperature, we see increasing median altitude. Meaning essentially that in a warm season, we're seeing a lot more malaria at higher altitudes than we're seeing in a colder season or cooler season. So in a sense, the malaria is climbing up in altitude if it's a mountain. The malaria climbs up the mountain as temperatures increase. Finally, these third graphs show year versus medium altitude and what we're looking at. So first versus mean the altitude which is in the blue. Median altitude and the red is showing mean temperature. So you can see that they track quite nicely. Maybe a little more nicely for Columbia than Ethiopia. So when we have a year with very high mean temperature, we also have a year very high meeting altitude. So finally, they created a model was a negative binomial regression model for those of you are familiar with these types of models and the variables that ended up being the best model are temperature, season, altitude, and interactions between these various variables. So that's consistent with the results that I showed you previously. So now let's turn to a second study. So this is a modeling study and it's looking at change in the length of the malaria season, sorry, malaria transmission season. We'll call it LTS. Comparing 2069 to 99 to 1980 to 2010 recording to an ensemble of five global climate models and five malaria models by RCP. The RCP represents the various scenarios for future emissions with the 8.5 being the business as usual scenario, which is the high emission scenario. Then the 2.6 scenario is the aggressive scenario where we take aggressive mitigation measures to limit emissions and then we have in-between scenarios. What are we seeing here? So we have the red being an increase in incidence between 2069 to 99 versus 1980 to 2010, and then we have this gradient. So blue means that there's a decrease in incidence. They also make it a little more complicated by taking into account the uncertainty of the models. So this pinkish color means that there's also predicting an increase but with more uncertainty. So what do we see here? So essentially these projections are consistent with the study that I showed you previously. We're seeing an increase really in highland areas. So that would include. So let's just focus on this is the RCP, 8.5 is the high emission scenario. Let let's focus on that one. So we're seeing increases in Eastern Africa showing the red, South Africa, Central Angola, Madagascar, Central America, Southern Brazil, and Eastern Australia. Those turns out are all highland areas. We're actually seeing some decreases in tropical regions, and that's probably due to the fact that hot drier climates are actually bad for the mosquitoes. So these models in this study that I've just been showing you had some limitations that didn't take non-climate drivers into account. So fairly large uncertainties. They didn't take into account which would really be impossible that parasite and or the vector may adapt to evolving climate conditions. So that would mean that the parameters used in the model would be invalid. Nevertheless, it's interesting that their projections made were very consistent with the study that I showed you as well as other studies that suggest that we're going to see more malaria at higher altitudes everything else being equal. So there are other models that do take non-climate drivers into account and they suggest that the spread distribution incidents of malaria will be determined primarily by not by the non-climate drivers. Those being economic development, urbanization, land use change, urbanization would actually tends to decrease malaria incidence as would economic development. Land use change, migration, and level of malaria control efforts which is probably the key factor frankly. So that being the key factor level of malaria control, that means that in a sense, it's in our hands, it's in the hands of good governance and political will. So with vigilance continuing progress in reducing malaria burden is probably possible despite climate change. That's number 1, the projection of the modest increase in length of transmission season and population at risk that I showed you in some locations doesn't take the non-climate factors into account. Climate change may be a primary driver. However, in specific locations which would be at the margins of the geographic distribution thinking again in terms of altitude at the margins where malaria tends to be endemic whereas it tends to not be present. So those are areas where particular vigilance has to be taken. Also, if there are endemic areas with poor control efforts, climate change might be particularly important in those areas.
