Here’s your next reading assignment. Read this short article on geostatistics by 8 a.m., Monday, February 20th. Be sure to login (using the link near the bottom of the sidebar) to the blog before leaving your answers in the comment section below.
- In this course, we’ve focused on the difference between a population and a sample drawn from that population. In the example in the article about rainfall in Sweden, what is the population and what is the sample?
- Suppose that in 1993 you selected a citizen of Belarus at random and tested the milk in his or her refrigerator for radiocesium. What are some reasons the level of radiocesium might vary from citizen to citizen? Which of these reasons would be relatively easy to model mathematically?
- What’s one question you have about the reading?
1. In that example, I think that the population would be all of the rain samples taken from the stations and the sample would be a test from a specific station.
2. There are many factors that could affect the radiation level of the milk including distance from the radiation leak, the location from where the cows lived, and other factors of the milk that affect radiation absorption. The distance from the radiation leak would be easy to model mathematically because as distance from the disaster site increased, radiation levels would be lower.
3. In the article, they say that Iodine prophylaxis is very effective and simple but Id like to know how effective it is compared to a normal thyroid with no radiation and with a thyroid with radiation.
1) The population would be the collective meteorological data for the rainfall (and wind direction) in Sweden. It is unfeasible to track rainfall all across Sweden and so the sample would consist of the data from the 700 monitoring stations that attempts to map out the rainfall in Sweden.
2) Possible reasons include the location of the citizen, wealth, and source of milk. The further away the citizen lives from the area of highest contamination, the less likely the milk would contain high levels of radiocesium and this would be easy to mathematically model. The wealthier the person, the more chances that he/she would be able to afford foreign, imported milk which would have passed quality control checks and would not be contaminated. The article mentions that citizens of Belarus living in villages do not have access to non-contaminated food and so wealth would be a major factor.
3) What are other examples of using filtered kriging?
1. The population is the entire area of Sweden, but the sample is the 700 meteorological stations where they collect rainfall across the country.
2. It might vary due to region, or age of the cow (are calves born after 1986 infected?). I think the easiest to measure would be region.
3. I’m confused on the concept of semivariogram. I think the idea is that latitude/longitude location doesn’t matter, just distance between two points. But I think they said semivariogram doesn’t apply in this situation. Is this because of the effect on geographic features on weather patterns?
1. The population is the entire meteorological data for the rainfall in Sweden, while the sample is the data collected from the 700 stations across Sweden.
2. From my point of view, the main factor is the region of the citizen, or the region of the farm which produced the milk. Another main factor is that the age of the cow. If the cow bore after 1987, then the it may suffer less effect from the radiation. However, the feeding(herbage) for the young cows may also be influenced.
3. What’s the advantage of Iodine prophylaxis?
1. In this case, the sample would be the set of observed rainfall values from the 700 meteorological monitoring stations across Sweden. The populatlion would be the theoretically infinite number of rainfall values in sweeden that could have been measured.
2. One reason for varying radiocesium contamination would be distance from Chernobyl itself. The GIS map of Belarus show that in general, the probability of milk contamination decreases as you get farther away from Chernobyl. This is a relatively easy factor to model mathematically. It gets more complicated when you take into account weather conditions which can be hard to predict. Another factor that complicates things is what dairy farms people get their milk from. For example, milk from an area close to Chernobyl might be shipped to an outlying region.
3. Is the map of Sweden showing the probability that rainfall was greater than 6 mm a prediction or a summary?
1) The population is the entire area of the rainfall in Sweden. The sample is the 700 monitoring stations.
2) Some reasons the level of radiocesium might vary from citizen to citizen could be where the citizen lives and where he or she gets the milk from. If the citizen lives farther away from Chernobyl, he or she is more likely to get milk with lower levels of radiocesium. If a citizen gets foreign milk for example, he or she would not get milk contaminated with radiocesium at all. But this would depend if a citizen can afford buying milk not from this region. I think that the place the citizen lives in and the level of contamination in the milk at this place would not be difficult to model.
3) Very informative reading; no questions.
1. The “population” would be the entirety of Sweden, over which rainfall as a continuous measurement would be impractical, so the sample was the rainfall measurements at each of the 700 meteorological stations.
2. The milk in a specific refrigerator could presumably come from any cow in the immediate region – While some would be born after the incident, meaning their only contamination would be through their grazing vegetation, the difference in radiocesium levels based on region would be fairly easy to model as a geographical heat map.
3. Are the semivariogram’s probabilities for un-mapped locations useful for relating to the important short-term weather patterns, or is the error high enough to render the semivariogram unhelpful?
1) The sample would be the soil area at the 700 measuring stations where rain occurred in the days following Chernobyl. The population is the total soil area in Sweden where rainfall occurred in those same days.
2) Where the person lived in Belarus (geographically), if their milk was bought commercially or obtained locally, and the diet of the cow in question. The easiest reason to model would be how the location of the person affected Cesium levels in the milk. This could be done easily by a heat map.
3) N/A
1) Population – all rain in Sweden coming from the direction of Chernobyl over the next few days after the incident
Sample – parts of the population measured by Swedish meteorological stations
2) Some citizens might import more food than others
Relative distance to Chernobyl blast (easy to model mathematically)
Certain food products might hold more radiation than others
3) How do geostatistics differ from “normal” statistics in terms of determining the probability of given events assuming they have these large, vaguely defined sample sizes?
1. In this course, we’ve focused on the difference between a population and a sample drawn from that population. In the example in the article about rainfall in Sweden, what is the population and what is the sample?
The population is the rainfall in Sweden and the sample is the tested rain
2. Suppose that in 1993 you selected a citizen of Belarus at random and tested the milk in his or her refrigerator for radiocesium. What are some reasons the level of radiocesium might vary from citizen to citizen? Which of these reasons would be relatively easy to model mathematically?
One that the article talked about is distance from the explosion, which was easy to model. Another that would not be as easy to model would be how much of the contaminated vegetables to the cows ate, and which ones since some vegetables may be more susceptible to contamination than others. .
3. What’s one question you have about the reading?
It said location was modeled as a formula of direction and distance, how did they mathematically model distance? With vectors or a number?
1. The population is all rainfall in Sweden, whereas the sample is the rainfall collected at the 700 meteorological monitoring stations.
2. The amount of radiocesium in the milk might vary depending on the make and model of the refrigerator, the age of the milk, what farm the milk came from, and where in Belarus the citizen lives. The location of the milk would be relatively easy to model mathematically.
3. While I felt like I could pick through the article to get details like the answers to questions 1 and 2, the big picture of the article was kind of unclear and I don’t think I understood what the article was trying to convey as a whole.
1. The population is all rainfall while the sample is the rain collected and measured at the 700 meteorological monitoring stations across the country.
2. The level of radiocesium might vary based on where the cows producing the milk are raised, what type of milk it is, where the citizen lives, and what the cows producing the milk are fed. Modeling the relationship between the level of radiocesium and where the citizen lives in relation to Chernobyl would be easy mathematically.
3. What is probability mapping, and how do you model it?
1) The population is the total rainfall in Sweden and the sample is the 700+ meteorological monitoring stations in Sweden.
2) The radiocesium contamination was spread out very nonuniformly across Sweden which would account for the difference. Also it would vary depending on the cow that produced the milk or even what time the cow produced the milk. The variance based off of geographic location would be the easiest to model mathematically.
3) What was the point of this if they didn’t even tell the residents of Belarus about the nuclear accident. It says that the point of predicting contamination based off of rainfall was so they could react faster than having the soil tested but then they don’t even tell the citizens.
1.
The sample was a combination of the regions which had one of the 700 weather monitoring stations. The population consisted of all regions of Sweden, and predictions were made for the unsampled regions.
2.
Some reasons may include geographic location, income, rainfall amount. It would be relatively easy to model the radiocesium amount in regards to geographic location.
3.
I’m confused about the assumption of stationarity. Could you explain that further?
1. The population is the rain in all areas of Sweden. The sample is Sweden’s meteorological network that consists of more than 700 meteorological monitoring stations
2. The level of radiocesium might vary because different citizens get their milk from different sources and may have different types of milk. Income may also affect what kind of milk people could afford to buy. It would be relatively easy to model income levels versus the level of radiation in a citizen’s milk.
1. The population is all the rainfall in Sweden around the time of the meltdown, the sample is everywhere it was measured.
2. Where they got the milk, and how far away from Chernobyl they lived, also weather patterns on the day of the meltdown. The last two would be fairly easy to mathematically represent.
3. none
A)
population : All fallen rain (which comes from Chernobyl direction) on every part of Sweden land after few days of the incident.
Samples: the measurable observations by the monitoring network. They do not cover all Sweden land accoording to the source (“In general, no matter how dense the monitoring network, there are many areas where observations are not available”)
B)
There are may factors. One of them is the kind of milk (noncontaminated or contaminated). Also, another factor is the awareness of the citizen about radiation. Third factor is meteorological data (we can not base it on actual distance.
Saying that, there are un-direct clear factors (which can be model mathematically). One of them is the wealth of the citizen. From that, we can know what kind food that citizen can offer (cheap food has more radiation than expensive ones (because of the noncontaminated operation and other operations). Another factor is citizen’s town. If it is a village, it is most likely to have more radiation because they make their own food (without any safety operations). However, if it is city, it is going to be less radiation because the offered food had gone via safety operation.
C)
What is iodine prophylaxis? (even after the article explains it, I still do not understand)
1.) The population are all the radioactive rain in Sweden, while the sample are the ones measured by then Swedish meteorologists.
2.)The level of radio-cesium might differ because of the difference in income and geography of the citizen. The income of each citizen would be relatively easy to model mathematically.
3.) None
1) Population – Europe, Sample – Sweden
2)The reasons may include where the milk is produced, and how much milk is consumed. It is easy to model the amount of milk consumed with the degree of the radiation given all consume the same type milk or milk produced at the same region.
3) No question.
1. Population: rainfall in Sweden
Sample: rainfall in Sweden in the areas around Chernobyl tested in April
2. Not all citizens get their milk from the same place (locally vs. non-locally) & not all citizens live close to Chernobyl. It would be relatively easy to model the location from where the milk was obtained vs the radiocesium concentration.
3. Why does the article include the amount of rain in millimeters? How is this relevant to the study?
1.
The population is all rainfall in Sweden after the Chernoybl disaster. The sample is the rain that falls where there are measuring stations.
2.
They may have different sources of their milk (ie from their own cows, local, commercially produced in the country, or even imported from a neighboring place). Their distance from the actual site would be an effective modelling statistic.
3. What’s one question you have about the reading?
Stationarity remained a fairly confusing term, and I am not sure what was implied by exact locations being more important than regions.
1. The population of the article is the total rainfall data from all over Sweden while the sample is the data that was recorded. Many areas of Sweden were not recorded in the data just because it would be impractical to take measurements of all the rainfall that occurred throughout the entire country.
2. The amount of radiocesium in the milk would depend on where the cows that produced the milk came from. Cows bred closer to the incident site are more likely to produce milk with higher levels of radiocesium. This distribution of cows that produce milk with higher levels of radiocesium would be fairly easy to model mathematically.
3. Will the effects of this radiation increase or decrease over time?
1) The population is all of Sweden and the sample is what areas they could actually measure and did not have to make a guess about.
2) The amount of radiocesium would vary quite a bit depending on where the cow that produced the milk was located. One could model this effectively by plotting the amount of radiocesium in the milk against the distance of the farm that the milk was produced at from Chernobyl. It would be expected that the greater the distance to Chernobyl the less radiocesium there would be
3) Why weren’t the people in Belarus warned that there was a high possibility that the food they consumed was hazardous to their health? Where was the government on this one? Seems like a pretty important thing to leave out.
1) The population would be all the people in Sweden within 10 years after the accident. The sample comes from those who are diagnosed with thyroid cancer within the same period.
2) Some citizens grow their own food and milk their own cows. If the citizen was within the affected area, that could be a reason radiocesium might be higher in their milk. Also, the level of contamination is dependent on the rain fall for a given area. Since the rain can fall more heavily in some areas than others, that is another reason for varying levels of contamination. It would be relatively easy to model the probability that the food of a person who grows their own produce is contaminated versus a person who buys their food from outside of the effected area.
3) The Kigring techniques seem interesting.
The population is the entire area of Sweden and the sample is the areas of the country with available meteorological data.
Most of the variation in the presence of radioactive material is due to the weather patterns in the days after the accident. Depending on where in Belarus the citizen is and where the cow is from could lead to very different data.
Do we need to know this?
1) Population – all the incoming rain from the direction of Chernobyl over the next few days after the incident in Sweden.
Sample – parts of the population measured by Swedish meteorological stations
2) Some citizens might purchase more food away from the Chernobyl explosion
Certain food products might hold more radiation than others
1. The population was a continuous geographic map of rainfall amounts. The sample consisted of the measurements at the 700 meteorological monitoring stations.
2. The radiation levels are determined by the location of the pastures where the cows grazed and what type of vegetation the cows eat. As well, soil type and weather patterns could further influence amount of radio-cesium retained and passed through cows to milk. The underlying factor in all of these is captured by the geographical location, and would be the easiest to model.
3. I’m not sure on what (or what the basic ideas of) kringing and a semivariogram are.
1) The population is all the rainfall in Sweden after the accident and the sample is the rainfall in areas that the scientists measured.
2) The amount of radio-cesium would vary in different citizen’s milk for many reasons. One being the citizen’s location within the city and where they do their grocery shopping. It would be easy to mathematically model where the citizens bought their milk and what farms that milk came from and its closeness to the blast.
3) Why is assuming stationarity so crucial in this article?
1) In this example, the population is the total rainfall in Sweden, and the sample is the individual points at which the rainfall data is taken.
2) Radiocesium contamination is distributed very non-uniformly, both geographically, and amongst different types of food. Probabilistic mapping of the geographical spread would be a good indication of which populations are at-risk.
1) The population in this case would be all the rain that fell in sweden after the incident and the sample would be the rain that was measured by swedish meteorologists.
2) Some citizens may get their milk at different grocery stores. Each of these stores may get their milk from a different area where the contaminations may vary from other areas. The area where the milk is obtained for each grocery store and the different grocery stores used by citizens may be mathematically modeled.
3) what else is GIS used for?
1. The population is all of the radioactive rain that falls in Sweden. The sample is the part of the populations that was measured by the Swedish meteorologists.
2. Reasons that the level of radiocesium will differ are how close the citizen lives in reference to the blast, what store the citizens buy their milk from and the the type of milk that they purchase. It would be easiest to model the proximity of the citizens to the blast site.
3. I wonder how Chernobyl is still affecting Belarus today?
1- The population is all the radioactive rainfall, and the sample is the measured rainfall.
2-The level of radio-cesium might differ because of the difference the distance from chernobyl that the cows were housed. If you knew the distance of the milk farms from chernobyl then you could model that mathematically.
3- None
1. The population is the rainfall information all over the area, and the sample is the rainfall observations collected by the 700 stations.
2. Possible reasons would be location and source of the milk. Location is much easier to model.
3. How can the food contamination not be distributed geographically?
The population is the land in sweden where it rained, and the sample is the places they had monitoring.
Some variance will come from origin of milk, whether it was produced locally or imported. Given that the milk is local, variance might arise from the distance of the producer from the reactor. This distance can be easily modeled mathematically.
Is all of the radioactivity only coming from the fallout, and are all increased cases of health effects caused by radioactivity, or could their be other factors, such as other chemicals that were introduced in local manufacturing in around the same period.
1. The population is all the locations which experience rainfall in Sweden; the sample is the approximately 700 meteorological monitoring stations located throughout Sweden. Data from the sample population was used to predict information about rainfall within the rest of the population.
2. The following are some reasons the level of radiocesium might vary from citizen to citizen:
– Genetics (some bodies may absorb the material more readily than others)
– Age (older age – less absorption may take place)
– Location of dairy farm the milk came from (and whether the cows in the farm were directly exposed to the rainfall or whether the cows lived indoors)
– Ingestion of a stable iodine pill (which would effectively prevent radioactive iodine adsorption)
Age and ingestion of a stable iodine pill would be relatively easy to model mathematically since the data is easily obtainable from individuals in the area – no complicated mathematics are involved in gathering information about age and pill consumption from people.
3. Clearly, it is important that the assumption of stationarity is met when using kriging. However, how do you know when this assumption fails? And when do you know that it is an entirely valid assumption to begin with?
1. The population is the people in the surrounding area, and the sample is those who went on record as having received cancer in the years following the nuclear disaster.
2. One reason may be the origination of that family’s milk. Another reason may be how much rain there has been in the area. The rain would be relatively easy to model, but whether or not the family purchased milk from a store that had its’ milk shipped in, or whether they got the milk from an independently owned farm would vary from house to house.
3. Was the rain supposedly caused by the nuclear fallout? Or were they trying to determine whether or not the rain fall realistically measured how much radioactive material was in the atmosphere? I was confused by this, and unfortunately, it seemed to be a main focus of the article.
1.) Population – all rain in Sweden from Chernobyl area
Sample – rain contamination measured by meteorological stations
2.) Origin of milk (import vs domestic or 2 different local areas), time of packaging vs rainfall in region. Any of these would be relatively easy to model.
3.) Would be interesting to know how the levels of radiation compare to the years prior to the meltdown.
1. sample: the amount of radioactive rain in 1986 in Sweden
pop: the amount of radioactive rain in Sweden over the next few years.
2. A few reasons could be the different kinds of food that each person eats, the different places that each persons food comes from, and amount each person ate. A mathematical model can be made of the amount of consumption by each person.
3. I Have no questions.
1. Population: All cases where threshold rainfall value was exceeded. Sample: Monitoring stations where threshold rainfall value was exceeded.
2. It might vary because of the source of milk, the time when the milk was “harvested”. Where the milk was sourced would be easy to model mathematically.
3. How does ArcGIS show uncertainty?
1. The sample is central and eastern Sweden, and the rain on 29 April that caused the radiocesium soil contamination. The population is all of Sweden and all of the rainfall that happened over those days.
2. The level of radiocesium might vary because the poorer citizens don’t have access to the noncontaminated food. In addition, to the winds and weather conditions across Belarus, the contamination would be distributed nonuniformly. It could be easy to model both of these by looking at income levels of various citizens and by geographically modeling the distribution of the contamination across Belarus.
3. For the most part, I found the article very inaccessible. For instance, the third paragraph in the “Radioactive Rain in Sweden” section is near-indecipherable without a large amount of prior knowledge outside of the scope of this class.
1.) The population is all areas in Sweden that could possibly have experienced radiated rain from Chernobyl. The sample is the locations that were measured, and not all locations that could have been affected were measured due to various difficulties.
2.) There should be a pattern in the amount of radiocesium in the milk farms distributed across Belarus, but the route the milk takes to the consumer is most likely completely independent of the distribution of radiocesium. That is, if there were a very heavy concentration of radiocesium in a milk production farm and the milk were to be shipped to many markets in Belarus, then there would be a relatively even distribution that would be much more uniform than the distribution of radiocesium due directly to Chernobyl.
3.) The description of Kriging is slightly vague. What is a more precise, but perhaps not very technical explanation of the process?
1. The stations collecting data on the rain are the sample. The rain over the entire country is the population.
2. Milk consumption could vary from person to person, or could be linked to availability of milk. It would be easy to model the level of radiation based on the amount of milk consumed.
3. Is the assumption that the rain was radioactive three days later a valid one?
1) The rainfall data gathered at the meterological testing sites are the sample data. The population is the all of rainfall that occured in Sweeden.
2) Better-off Belarusians likely have access to safer food products. It would be easy to mathematically model that. Other reasons for why the level of radiocesium would vary from citizen to citizen could be geographic in nature, because the initial distribution of radiocesium was non-uniform. Some of those reasons would be harder to model than others. For example, it might be relatively easy to model the probability that milk would be contaminated given the direction of the wind during the days after the disaster.
3) How is kriging conducted?
1. In the article about rainfall in Sweden, the sample is the data collected by Sweden’s meteorological network of more than 700 meteorological monitoring stations. This data is then used to predict the behavior caused by rainfall in areas that are unsampled. The population is the total rainfall in all of Sweden including the unsampled areas. This is predicted by the semivariogram to quantify the spatial correlation in the data.
2. The levels of radiocesium in milk in 1993 in Belarus would vary from citizen to citizen because the amount of radiocesium spread by the Chernobyl incident varies. This variation is dependent on the distance from the Chernobyl plant and the weather patterns which carried the radiation from the plant. The variation that is easiest to model mathematically would be the distance factor, because the amount of radiation found would decrease as you get further away from the plant.
3. What is a semivariogram and how is it made?
1) The population is the complete data of all the rainfall over the entire country. This includes all territories. The sample is where the meteorological monitoring stations were set up. These were set up at particular areas.
2) Radio-cesium is distributed non-uniformly geographically and within different types of food. Some cows could eat more than others consuming more radioactive chemicals. The level of consumption of particular cattle would be easy to model mathematically because all cows eat the same stuff.
3) What comparison do they draw their high levels from? It seems they just measure rainfall and assume the rain is contaminated.
1. The population is Sweden, and the sample is the group of monitoring stations.
2. The level might depend on the consumption of outside, imported food, which can reasonably be modeled by income level.
3. What is “Kriging”? I didn’t quite understand.
1. the sample is the rainfall data. population is the rainfall data in all the territories.
2. radiocesium concentration is distributed nonuniformaly geographically and with different types of food. they can both be presented mathematically with probability mapping.
The population was rain and the sample was rain after the radioactive incident.
One reason for the difference in radiocesium levels in milk would be income. Families with higher income would be able to buy milk from places unaffected by the fallout, while poorer families would have to buy local milk or even raise their own cows for milk.
1. The population was the areas (as discrete as that is) in Sweden to collect rainwater from. The sample was the data taken. I do not think the ‘assumed’ or calculated areas would count in the sample. Right?
2. The location could be a big deal! Winds could have taken the radiocesium to some locations much more than others. Maybe it didn’t rain in some locations too. The milk could’ve even come from a different place.
3. Nothing really, would just like to know more about the calculations they did to predict the areas they did not test in the sample!
1. The population is all the rain in Sweden a few days after Chernobyl. The population is the data from the meteorological centers.
2. Where the person is located, where they bought there milk. Where they live would be easy to model.
3. What does “aggregated over polygonal regions” mean?
1. The population was likely radioactive rainfall over all of Sweden. The sample taken was likely radioactive rainfall in central and eastern Sweden.
2. One reason is the citizen’s proximity to Chernobyl, which would affect the total amount of radiation nearby cows would be getting. This would be easy to mathematically model, as distance from the explosion would generally mean less radiocesium in the milk. Another reason could be where the citizens got their milk from. This would be more painstaking to model since it would take a lot of data collecting from individual citizens.
3. How would we figure out how much more important distance from Chernobyl is as a variable when compared to amount of rainfall in an area (or if it’s less important)?
In this course, we’ve focused on the difference between a population and a sample drawn from that population. In the example in the article about rainfall in Sweden, what is the population and what is the sample?
The article was all over the place; samples were taken from Sweden, Belarus, Mink etc. If I had to summarize, I’d say the sample is the distinct individuals who took part in the data provided by the Sakharov Institute of Radioecology, Minsk, Belarus.
and the population is all current individuals at risk of radioactivity exposure from Chernobyl.
Suppose that in 1993 you selected a citizen of Belarus at random and tested the milk in his or her refrigerator for radiocesium. What are some reasons the level of radiocesium might vary from citizen to citizen? Which of these reasons would be relatively easy to model mathematically?
Where the milk was from. Whether it was shipped from areas close to Chernobyl or not. How old the milk is. Whether it was cow’s milk or not. The cow’s diet (whether the cow ate radioactive feed). Where the milk was stored.
It would be easy to model most of these. For example, to model the cow’s diet, you could graph radioactivity of the milk vs radioactivity of the cow’s diet to determine a positive association.
What’s one question you have about the reading?
I had quite a bit of difficulty following parts of this reading. What is GIS and spatial statistics, disjunctive kriging, and I really had no idea what semivariogram meant.
In this course, we’ve focused on the difference between a population and a sample drawn from that population. In the example in the article about rainfall in Sweden, what is the population and what is the sample?
The article was all over the place; samples were taken from Sweden, Belarus, Mink etc. If I had to summarize, I’d say the sample is the distinct individuals who took part in the data provided by the Sakharov Institute of Radioecology, Minsk, Belarus.
and the population is all current individuals at risk of radioactivity exposure from Chernobyl.
Suppose that in 1993 you selected a citizen of Belarus at random and tested the milk in his or her refrigerator for radiocesium. What are some reasons the level of radiocesium might vary from citizen to citizen? Which of these reasons would be relatively easy to model mathematically?
Where the milk was from. Whether it was shipped from areas close to Chernobyl or not. How old the milk is. Whether it was cow’s milk or not. The cow’s diet (whether the cow ate radioactive feed). Where the milk was stored.
It would be easy to model most of these. For example, to model the cow’s diet, you could graph radioactivity of the milk vs radioactivity of the cow’s diet to determine a positive association.
What’s one question you have about the reading?
I had quite a bit of difficulty following parts of this reading. What is GIS and spatial statistics, disjunctive kriging, and I really had no idea what semivariogram meant.
1. Population is all rainfall in Sweden. Sample is radioactive rain that was measured.
2. Depends on where you buy the milk.
3. Do we use semivariograms a lot?
1. The population is all rain in Sweden immediately following Chernobyl. The sample is that rain in the population that could be measured by meteorology stations across the country.
2. imported milk or milk from cows brought into the country well after the Chernobyl incident. This could be tracked through the trade patterns and data such as GDP categorical breakdowns for Belarus
3. None
1. Population is all the rainfall in Sweden. Sample is the rainfall data taken from 700 meteorological monitoring stations in Sweden .
2. The reasons that the radiocesium level varies from citizen to citizen are incomes and the source of their diet. Income would be easy to model mathematically.
3. none
2. The differences in contamination could be as result of a number of factors such as distance from the original site, wind speed, amount of rainfall received since the initial explosion, etc. among all these the easiest to model would be the distance as it is very easy calculable. The other 2 factors will require special measuring equipment which may not be available in all places.
3. I found no specific reference to data in the Swedish rainfall paragraph. What exactly was indicative of population or samples in the article?
1. Population: People under age 15 in Sweden. Sample: Those from this experiments population who had thyroid cancer (per year).
2. Radiocesium varies geographically because different amounts of radiation are present in different parts of the country. If two individuals got their milk from two dairy farms located in different parts of the country, then radiocesium would also differ most likely. Other factors might include type of milk and how it interacts with the radiocesium, or how the weather on the day of bottling the milk might effect the concentrations of radiocesium. The first is by far the easiest to model.
3. Would those charts with different sized bubbles representing populations for various countries be considered geospatial modeling in any way?
1. The population is the amount of radionuclide contamination in the ground. The sample is the meteorological data (rainfall).
2. It may depend on the city’s distance to the source point: Chernobyl, on the direction of rainfall after the incident, and whether or not the citizen was aware (in advance) to take Iodine prophylaxis after the accident. Using the meteorological data, I think the direction of rainfall can be mathematically modeled.
3. The histogram data showed that the clear increase in reported thyroid cancer did not happen until about 4 years after the accident. Is this because it is a slowly reacting radiation or is because it needs to build up to a certain amount for it to cause cancer?
1.Population: All of Sweden, Sample: 700 meteorological monitoring stations in Sweden
2.They might reside in different parts of the country where different levels of radiocesium were present; the milk might be different brands, produced in different parts of the country, where different levels of radiocesium were present, etc. It would be relatively easy to make a map of general radiocesium concentration throughout the country.
3.Why was milk in particular contaminated? How can the article say that milk accounted for 36% of the internal radiocesium dose in rural populations? How can the authors be certain of which foods were affected?
1) population = every raindrop on Sweden
sample= the rain measured by the 700 meteorological stations
2) Yes, a lurking variable would be the expiration date of the milk, which could be measured by looking on the bottle
3) what is another example of a semivariogram?