Filename
stringlengths 22
64
| Paragraph
stringlengths 8
5.57k
|
|---|---|
Processed_Gauß_Cubature_for_the_Surface_of_the_Unit_Sphere.txt | Lebedev [14] published an important paper on Gauß cubature for U3 just prior to that of the more general one by Mantel and Rabinowitz [18]. Lebedev makes some astute choices in the presupposition of rule structures, which sometimes result in optimal choices. The algebra is simplified by enforcing K1 = K2 = K4 = 1, and sometimes also K2 = 0; and choosing K3, K4, and K5 such that the number of unknowns in ( ) is equal to the number of equations. The rules generated were a ∗ great improvement on the previously completely unsystematised collection of known rules (e.g. see Stroud [26]), but we are interested in the more general case. |
Processed_Gauß_Cubature_for_the_Surface_of_the_Unit_Sphere.txt | Table 5: Basic solutions to the IPP, with the integer constraint removed. Any integer solutions will have N at least equal to that displayed here. |
Processed_Gauß_Cubature_for_the_Surface_of_the_Unit_Sphere.txt | We require all the solutions to the IPP which have N at least equal to the (integer) minimum, and less than some (user-specified) small multiple of this minimum. We firstly solve the programming problem without integer constraints. This yields a lower bound for N , used when searching for all the integer solutions. This initial problem is solved using the Mathematica program IPPBasicSoln.M (Appendix E). Results (lower bounds for N , and associated ‘pseudostructures’ ), for m = 1, . . . , 20 are presented in Table 5. |
Processed_Gauß_Cubature_for_the_Surface_of_the_Unit_Sphere.txt | To find all the solutions of the IPP, we use a simple exhaustive search. Whilst this is crude, in this case it is reasonably efficient. For U3, there are only six variables . We use the lower bound on N , and as an to search through, and K1, K2, K4 ∈ { } upper bound we use a small multiple (say 1.5) of the lower bound. This second phase is implemented as a C program ipp.c (Appendix C). This program takes the data in Table 5, and exhaustively searches for all integer solutions, subject to (reasonable) bounds K3, K5, K6 6 20, and an upper bound on N set empirically so that we only collect about the first 100 solutions. The output from ipp.c is (manually) sorted, so as to order the solutions firstly with increasing N , and then lexically. The solutions corresponding to the first five integer minima, for m = 1, . . . , 10 are presented in Table 6. |
Processed_Gauß_Cubature_for_the_Surface_of_the_Unit_Sphere.txt | When ( ) has multiple solutions, the labelling is not unique, but this is not a ∗ problem. This nomenclature provides a basis for comparison with (published) rules derived by other means. |
Processed_Gauß_Cubature_for_the_Surface_of_the_Unit_Sphere.txt | These moments are calculated by U3Moments.M, a Mathematica program (Ap- pendix E). Table 7 lists them, for m 6 10, and this allows us to establish ( ) for ∗ Gauß rules of degrees 3, 5, . . . , 21. |
Processed_Gauß_Cubature_for_the_Surface_of_the_Unit_Sphere.txt | Having constructed tables of moments and possible structures for various m, we may consider solution of (the pared-down version of) ( ) for U3. Attempting to ∗ use Mathematica to do this succeeded for m up to 5 (sometimes with a little human assistance in making substitutions). Beyond m = 5, the high degree of the ) in general has only transcendental solutions. polynomials involved means that ( ∗ ) is attempted using numerical software in matlab (and Approximate solution of ( ∗ C). |
Processed_Gauß_Cubature_for_the_Surface_of_the_Unit_Sphere.txt | We use the matlab routine fsolve, which numerically approximates the solu- tion of a system of equations. This routine requires a user-specified function that ). Initially, this was done as a matlab m-file. As the size of evaluates the system ( ∗ ( ) increases rapidly with m, the large number of expensive function evaluations ∗ involved made this a slow procedure. Instead, momenteq.c, a matlab mex-file was written (Appendix D), which yielded a speedup of two orders of magnitude. As numerous experiments are required to find a satisfactory structure, this speedup is important. To aid debugging, a program writestar.c (Appendix C), takes in- puts of m and ) as a set of displayed equations. |
Processed_Gauß_Cubature_for_the_Surface_of_the_Unit_Sphere.txt | A matlab driver program, cubature.m (Appendix D) is used to try various structures and choices of m. For each m, cubature.m is run until either fsolve solves ( ), or the user gives up in disgust. Successful results for m = 1, . . . , 8 are ∗ presented in Appendix B. |
Processed_Gauß_Cubature_for_the_Surface_of_the_Unit_Sphere.txt | Product rules are an important approach to multidimensional cubature, and they are the main direct competitor with Gauß rules. Here we describe their construc- 3. The tion for U3, and compare their efficiency with the Gauß rules discussed in material is largely abstracted from chapter 2 of Stroud [26]. |
Processed_Gauß_Cubature_for_the_Surface_of_the_Unit_Sphere.txt | In fact [18] shows that in theory, an FSM rule for C3 can be constructed using 345 points, although there is no explicit calculation of such. It seems likely that an FSGM rule on less than 400 points exists. This book [26] illustrates the limitations on the usefulness of product rules in terms of the relationship N = mn. Table 8 illustrates some upper limits on m for various n if a ceiling of N function evaluations is enforced. |
Processed_Gauß_Cubature_for_the_Surface_of_the_Unit_Sphere.txt | This is not expressible in a simple closed form, but calculation is routine and efficient (e.g. the implementation in gauss.m in Appendix D). |
Processed_Gauß_Cubature_for_the_Surface_of_the_Unit_Sphere.txt | xi, Bi − { xi = (νi,1, νi,2, νi,3)⊤ and Bi = A1,i1 A2,i2 . |
Processed_Gauß_Cubature_for_the_Surface_of_the_Unit_Sphere.txt | Some substitutions and relabelling in the construction shows that it is actually simple, and we write the computations as an algorithm. |
Processed_Gauß_Cubature_for_the_Surface_of_the_Unit_Sphere.txt | Implementation as two matlab files is presented in Appendix D. Specific illus- tration of the points and weights of the rules generated is rather pointless, but a table of errors for integrating the appropriate polynomials generated by the function u3prod.m shows it to work perfectly (the errors are of order machine precision). |
Processed_Gauß_Cubature_for_the_Surface_of_the_Unit_Sphere.txt | Table 9: Comparison of the number of points required for the FSGM (or FSMG) Gauß rules and the product rules for U3. The last column is the ratio of the number of points (function evaluations) required by the Gauß rules relative to the product rules. |
Processed_Gauß_Cubature_for_the_Surface_of_the_Unit_Sphere.txt | Gauß rules (and their ancestors the equally-spaced formulae) are based on exploiting the analytical properties of smooth integrands. They are optimal in the sense that in general they will be the best choice for approximating integrals involving smooth integrands. As mentioned in 2.2, the optimal multidimensional cubature can only be sensibly considered for regions with some symmetry, and we dealt with the case of full symmetry. |
Processed_Gauß_Cubature_for_the_Surface_of_the_Unit_Sphere.txt | To deal with non-smooth integrands, possibly even random distributions, and with non-symmetric, possibly even disconnected domains of integration, alternative philosophies are usually more relevant. Textbooks on numerical integration com- monly contain many pages describing minute implementational details to further refine the theory for optimal methods; and then devote a similar amount of space to describing real alternatives (e.g. [6, 26]). The main alternative technique is the Monte Carlo method, but a more recent idea is the ‘lattice’ method. |
Processed_Gauß_Cubature_for_the_Surface_of_the_Unit_Sphere.txt | These techniques are based on averaging function values at a random selection of points within the region, and they are particularly appropriate for oddly-shaped regions and non-smooth integrands. They are widely used in statistical applications, where integrals of high dimension must be approximated. Performance is often about N −1 (the error incurred using N points should be proportional to N −1). Volumes have been written about them (e.g. see [26, chapter 6]). |
Processed_Gauß_Cubature_for_the_Surface_of_the_Unit_Sphere.txt | The problem with attempting to generalise this to the case of several dimensions is that the notion of ‘equal spacing’ of evaluation points becomes less well-defined. Placement of equally-spaced points on U3 is equivalent to maximally covering it with nonintersecting equal circles, a problem thought to be intractable. The ‘best’ that can be done involves heuristic algorithms, and lots of computer time [3, 4]. |
Processed_Gauß_Cubature_for_the_Surface_of_the_Unit_Sphere.txt | Nevertheless, it may be reasonable to try to approximate the equally-spaced placement of points within our region. Recent research involving Sloan and Lyness [16, 21, 22], has achieved this using geometric construction techniques, called ‘lattice methods’. For periodic functions on [0, 1]N , lattice methods generalise the trape- zoidal rule, preserving the order of the error as N −2. For U3, a placement called ‘spherical t-designs’ is used [21]. |
Processed_Gauß_Cubature_for_the_Surface_of_the_Unit_Sphere.txt | For FS regions and weight functions, where the expected integrands are smooth, these alternatives are a poor second choice in comparison with Gauß rules. The Monte Carlo methods will only converge as N −1, and the theory for the lattice methods is not very general, results only being available for one type of region at a time. |
Processed_Gauß_Cubature_for_the_Surface_of_the_Unit_Sphere.txt | It cannot be overemphasised that where Gauß cubature is available, it should be used, particularly as the dimensionality increases. For evaluation of integrals over 2D manifolds, Gauß cubature is applicable. |
Processed_Gauß_Cubature_for_the_Surface_of_the_Unit_Sphere.txt | Results from running findec.c (Appendix C) for n = 1, . . . , 100 are listed in Table 10, and graphically presented in Figure 4. For n = 100, the (optimised) program requires about 1000 CPU minutes on a SPARC-10 workstation. Data has been manually checked for n = 1, . . . , 10. |
Processed_Gauß_Cubature_for_the_Surface_of_the_Unit_Sphere.txt | Table 10: The number e + 1 of types of equivalence classes of FS sets of points for n = 1, . . . , 100. |
Processed_Gauß_Cubature_for_the_Surface_of_the_Unit_Sphere.txt | We present computed FS (usually FSMG) Gauß rules for U3, for m = 1, . . . , 8 (de- grees 3, 5, . . . , 17). For m 6 5, Mathematica (or a combination of it and some manual substitutions) provides analytic solutions. For higher degrees, only tran- scendental solutions exist. Instead, we use the matlab program cubature.m to approximate a solution, and we have to trust that an approximation with a small residual corresponds to a transcendental solution. The analytic solutions for low degrees provide good test data for cubature.m. All of the rules presented have been discovered by cubature.m; where possible, analytic solutions have been substituted. Beyond m = 8, no solutions at all have been found, but they should be available with sufficient computational effort. For most cases the rules are FSMG from the first structure corresponding to the first minima of the IPP. For m = 4, the struc- ture is from the second minima of the first structure. For m = 6 there is no FSMG rule, but an FSGM rule from the first minima of the second structure is obtained. |
Processed_Gauß_Cubature_for_the_Surface_of_the_Unit_Sphere.txt | Find the number of types of equivalence classes e+1, of fully symmetric points in n dimensions, by exhaustively enumerating them. A UNIX input line "findEC a b" will find e+1 for n = a, ..., b. |
Processed_Gauß_Cubature_for_the_Surface_of_the_Unit_Sphere.txt | Exhaustively solve the IPP, using the precalculated lower bound for $ N $. This is a crude but fast and successful method. The data is output, for each m, in terms of increasing structure. Running the UNIX "sort" on the output orders it into increasing N, then structure. |
Processed_Gauß_Cubature_for_the_Surface_of_the_Unit_Sphere.txt | C code that compiles into a MATLAB .mex file that ‘evaluates’ the moment equations (*). |
Processed_Gauß_Cubature_for_the_Surface_of_the_Unit_Sphere.txt | if (j+k+l <= t) i = [j k l]; tabi(m,:) = i; table(m) = 2*(y(:,1).^(2*i(1)).*y(:,2).^(2*i(2)) ... |
Processed_Gauß_Cubature_for_the_Surface_of_the_Unit_Sphere.txt | Find basic solutions to the IPP, ignoring the integer requirement, for m = 1, ..., 20. |
Processed_Gauß_Cubature_for_the_Surface_of_the_Unit_Sphere.txt | Interpolation and hyperinterpolation. Technical Report 91 – 22, Mathematical Sciences Institute, Cornell University, May 1991. Applied Mathematics Preprint AM91/16. |
Processed_Should_international_borders_re-open?_The_impact_o.txt | Novel coronavirus disease (COVID-19) has spread across the world at an unprecedented pace, reaching over 200 countries and territories in less than three months. In response, many governments denied entry to travellers arriving from various countries affected by the virus. While several industries continue to experience economic losses due to the imposed interventions, it is unclear whether the different travel restrictions were successful in reducing COVID-19 importations. Here we develop a comprehensive framework to model daily COVID-19 importations, considering different travel bans. We quantify the temporal effects of the restrictions and elucidate the relationship between incidence rates in other countries, travel flows and the expected number of importations into the country under investigation. As a cases study, we evaluate the travel bans enforced by the Australian government. We find that international travel bans in Australia lowered COVID-19 importations by 87.68% (83.39 - 91.35) between January and June 2020. The presented framework can further be used to gain insights into how many importations to expect should borders re-open. Authorities may consider the presented information when planning a phased re-opening of international borders. |
Processed_Should_international_borders_re-open?_The_impact_o.txt | On 11th March the World Health Organisation (WHO) declared the outbreak of novel coronavirus disease (COVID-19) a pandemic1. By then over 110,000 cases of COVID-19 had been confirmed in 114 countries and territories1. The virus with its origin in the Chinese city of Wuhan2, was quickly introduced into other regions and countries through international travel. As a consequence, governments around the world enforced travel restrictions and closed international borders, which resulted in a sharp decrease of passenger flights3. While the negative economic effects of the interventions are clearly visible and can for example be measured through financial performance indicators and an increase in unemployment, the societal benefit of restricting international travel has not been quantified. With the growth of outbreaks beginning to slow in some countries, calls to re-open borders are increasing. To assess whether it is feasible to do so, it is important to understand the risks posed by those countries that could act as importation sources. |
Processed_Should_international_borders_re-open?_The_impact_o.txt | There are two major factors that influence the expected number of COVID-19 importations: incoming traveller volumes, and incidence rates of the disease in source countries. In Australia, the border force collects information about every individual entering the country, which is published in anonymised and aggregated form. To assess the travel bans implemented by the Australian government, we estimate passenger volumes for the hypothetical scenario that no travel restrictions were implemented, using seasonal auto-regressive integrated moving average (SARIMA) models4. |
Processed_Should_international_borders_re-open?_The_impact_o.txt | Country-level COVID-19 incidence data is available on the website of the European Centre for Disease Prevention and Control (ECDC). There is consensus that the ascertainment of COVID-19 cases is extremely low, with some studies claiming that less than 1% of cases are reported to authorities5–8. Hence, the available incidence data very likely does not reflect the true scale of the pandemic and needs to be adjusted to account for this bias. We perform maximum-likelihood estimations using the observed incidence of the disease in travellers arriving into Australia to adjust the data. Similarly, to adjust for under-reporting within Australia we perform maximum-likelihood estimations using the observed incidence in Australian travellers arriving into New Zealand. |
Processed_Should_international_borders_re-open?_The_impact_o.txt | overseas as well as possible in-flight transmission. The model presents a flexible framework that can be used to quantify the effects of travel restrictions and to evaluate proposed relaxations. Further, it allows the identification of groups of travellers that most likely carry the virus, which can inform strategies for the optimal use of available prevention and control resources. In contrast to other studies that have looked at exportation risk from China9–11, our model quantifies the expected number of importations for a particular country and is global in scale. |
Processed_Should_international_borders_re-open?_The_impact_o.txt | We use Australia, one of the first countries to report imported cases of the disease12, as a case study to demonstrate the model’s capabilities. As an island country, nearly all international travel into the country is via air, which allows tighter control of international borders. Although partial border closures came into effect as early as one week after the first case was detected on 25th January, Australia experienced an exponential growth in the number of reported imported COVID-19 cases until the end of March13. This study sheds light onto the effectiveness and timeliness of the individual travel bans implemented by the Australian government. |
Processed_Should_international_borders_re-open?_The_impact_o.txt | Importations under different travel bans We model the expected number of COVID-19 importations into Australia between 1st January and 30th June 2020. In particular, we consider two different scenarios. The first considers the actual travel restrictions as implemented by the Australian government. The second scenario, which we refer to as open borders, is hypothetical and assumes that no travel restrictions were enforced. |
Processed_Should_international_borders_re-open?_The_impact_o.txt | Figure 1. Estimated daily COVID-19 importations. Our model estimates that a total of 6,003 COVID-19 cases were importations into Australia between 1st January and 30th May 2020, considering the current travel restrictions (solid line). Without any travel restrictions a total of 48,715 cases would have been imported during the same time period (dashed line). The shaded are indicated the 95% confidence interval of our estimations that was obtained by averaging over 100 model runs. |
Processed_Should_international_borders_re-open?_The_impact_o.txt | Figure 1 compares the expected number of importations for the two scenarios. The first travel ban that was implemented by the Australian government affected foreign nationals travelling from, and transiting through China within the last 14 days prior to arrival14. Figure 1 shows a clear reduction in the expected number of importations after this restriction came into effect on 1st February 2020, compared to the open border scenario. The difference in the expected number of importations for the two scenarios becomes smaller towards the end of the month, which is likely due to China being successful in containing the outbreak. Our results indicate that Australia was able to lower COVID-19 importations from China by 94.45% (91.77 - 96.32) during the studied period. We estimate that approximately 1,938 fewer cases were imported from China, out of the total 2,052 importations projected for that period with open borders. The arrival card data reveals that the reduction of importations cannot solely be attributed to the travel ban itself, which only affected foreign nationals. During February only 19.57% of the expected number of Australian citizens/residents returned from China, thus, contributing to the reduction in importations. |
Processed_Should_international_borders_re-open?_The_impact_o.txt | The second travel ban affected foreign nationals arriving from Iran and came into effect on 1st March 202015. Our results indicate that this restriction was less effective than the travel ban on foreign nationals arriving from China. COVID-19 importations from Iran were reduced by 32.81% (0 - 56.88). Overall, only 14 fewer cases were imported from Iran. Presumably, this travel ban was less effective due to the majority of arrivals from Iran being Australian citizens and residents who were exempt from the ban. |
Processed_Should_international_borders_re-open?_The_impact_o.txt | On 5th March 2020 Australia denied entry to all foreign nationals arriving from South Korea16, which resulted in a 94.41% (92.05 - 96.14) reduction of cases being imported from this source. In contrast to arrivals from Iran, arrivals from South Korea are dominated by foreign travellers, explaining the high reduction in importations from this country. In addition, only 5.49% of the expected number of citizens/resident arrivals returned to Australia during March. Overall, we estimate that 433 fewer cases were imported from South Korea. |
Processed_Should_international_borders_re-open?_The_impact_o.txt | Six days later, on 11th March 2020 foreign nationals arriving from Italy were banned from entering Australia17. This travel ban reduced the number of COVID-19 importations from Italy by 77.9% (69.21 - 85.76). In total, 994 fewer cases were imported. However, only 36.33% of the expected number of citizens/resident arrivals returned to Australia. |
Processed_Should_international_borders_re-open?_The_impact_o.txt | The final travel ban enforced by the Australian government denied entry to all foreign travellers. This restriction was implemented on 20th March18. Figure 1 shows a sharp decrease in the number of importations on this date. We estimate that Australia imported on average between 15 and 22 cases a day between 21st March and 30th April. During May and June the daily average dropped to three cases a day. Our results show that the reduction of COVID-19 importations is partly due to fewer Australian citizens and residents returning than expected during non-pandemic conditions. A significant factor underlying this reduction is in reduced flight availability into the country. The reduction of importations directly attributable to the individual travel bans is discussed later in this article. |
Processed_Should_international_borders_re-open?_The_impact_o.txt | Figure 2. Comparison of importations by citizens/residents and visitors before and after the implementation of travel bans. The stacked bar chart shows the estimated number of importations by Australian residents and citizens before (dark blue) and after (dark green) the date of the travel ban. The light blue and light green bars show the estimated importations by visitors before and after the date of the travel ban, respectively. |
Processed_Should_international_borders_re-open?_The_impact_o.txt | was the United States of America with an estimated 957 (903, 1,036) cases. A full ranking of importation sources is provided in Supplementary Data File 1. To better evaluate the individual travel bans, we distinguish between foreign nationals visiting Australia and citizens/residents of Australia who were exempt from all restrictions. We note that under certain circumstances visitors were allowed to enter the country after the establishment of travel bans, for instance, if they are immediate family members of a citizen or resident. |
Processed_Should_international_borders_re-open?_The_impact_o.txt | Figure 3. Estimated percentage reduction of imported COVID-19 cases. The estimated cumulative number of importations by Australian citizens/residents (dashed curve) and visitors (solid curve), assuming no travel bans are implemented. The vertical dashed line indicates the date when the cumulative number of visitor importations reached one. The corresponding label shows the expected percentage reduction in the total number of importations over the studied period if a travel ban had been implemented on the same day. The solid vertical line indicates the implementation date of the actual travel ban and the corresponding percentage reduction in imported cases. |
Processed_Should_international_borders_re-open?_The_impact_o.txt | Figure 2 shows the estimated number of COVID-19 cases imported by citizens/residents and visitors from the ten largest sources before and after the respective travel restrictions were implemented. In most cases more COVID-19 cases were imported before the implementation of travel bans with Italy being the exception. The ban on foreign travellers from Italy came into effect on 11th March 2020. At the same time the government urged citizens and residents to return to Australia as soon as possible. The border force recorded an average of 130 citizen/resident arrivals from Italy each day until the end of March when the outbreak in Italy peaked. The estimated increase in visitor importations from Italy is likely due to the increased return of immediate family members of returning citizens/residents who are not citizens or residents themselves. |
Processed_Should_international_borders_re-open?_The_impact_o.txt | Effectiveness and timing of travel bans We showed that the overall reduction of COVID-19 importations is not only due to the individual travel restrictions, but can partly be attributed to fewer Australian citizens/residents returning from overseas. In this section, we quantify the direct effect of the individual travel restrictions. To do so, we assume that travel bans hinder all visitors who arrive from the corresponding country from entering Australia, as was intended by the government. In addition, we assume that all citizens/residents continue to return to Australia as usual. Our calculations are based on the total number of projected arrivals from each source country during the study period. |
Processed_Should_international_borders_re-open?_The_impact_o.txt | Figure 3 displays the results for the nine largest sources of visitor importations. The dashed and solid blue curves show the estimated cumulative number of importations by citizens/residents and visitors for the open borders scenario, respectively. The dashed vertical line indicates the date of the first importation from a specific source country. Its label shows the percentage reduction in total importations over the studied period that could have been achieved if visitors were banned from this date onward. The solid vertical line indicates the actual implementation date of the travel ban and the respective percentage reduction in imported cases. |
Processed_Should_international_borders_re-open?_The_impact_o.txt | three respective sources. Among the studied countries, the reduction in importations from Austria was the lowest. Only 66.26% of preventable importations could be averted. The extent of importation reductions (solid vertical lines) are determined by the incidence rate in source countries, the travel volume from that country, and the number of days after the day of first importation from that country. Implementing travel bans closer to the date of first importation can further reduce the importations from a source country. |
Processed_Should_international_borders_re-open?_The_impact_o.txt | Devising the re-opening of borders To decide whether it is safe to open international borders, governments need to understand the relationship between the number of arrivals, incidence rates in countries that act as importation sources and the expected number of COVID-19 importations. The contour plot in Figure 4 visualises this relationship. We assume that arrivals spent an average of 15 days in the source country prior to arrival. Stars indicate the expected number of importations during October 2020 from the corresponding country if borders re-open and given the country’s current reported incidence rates and the expected number of arrivals. The countries referenced in Figure 4 were amongst the most frequently cited origins of travellers arriving into Australia in 2019. Note that the US, UK and Indonesia are not shown in Figure 4 as the recent daily incidence rates are well above 1e-6. |
Processed_Should_international_borders_re-open?_The_impact_o.txt | Figure 4. The relationship between daily incidence rates, the number of arrivals and the expected number of importations. We assume that arrivals spent an average of 15 days in the source country. Darker areas of the contour plot indicate a higher number of expected importations. Stars mark the expected number of importations from several countries during August 2020 if borders were to re-open. The results are averaged over 1,000 model runs. |
Processed_Should_international_borders_re-open?_The_impact_o.txt | Our results indicate that Australia can expect less than one importation from Malaysia, China, Fiji and Thailand during October if borders were to re-open. However, there is still a risk of COVID-19 cases being imported. Assuming that the number of infected arrivals can be modelled by a Poisson process with rate parameter equal to the expected number of importations, the probabilities of importing one or more cases from countries where we expect less than one importation during August, are as follows: China: 4.33%, Fiji: 12.78%, Malaysia: 56.87%, Thailand: 5.66%. |
Processed_Should_international_borders_re-open?_The_impact_o.txt | We developed a comprehensive framework that models importations of COVID-19 into any given country. The model can be used to quantitatively evaluate existing and proposed border closures and is useful in guiding authorities to decide whether current travel restriction can be relaxed. |
Processed_Should_international_borders_re-open?_The_impact_o.txt | We applied our model in the context of Australia, yet the model is globally applicable to any jurisdiction that tracks the movement of people through its borders. Mapping incidence rates and arrivals from source countries to the expected number of importations can underpin a decision support tool to determine the country-specific risk of opening international borders. It is important to note that COVID-19 reporting mechanisms and protocols can vary greatly across countries, and these practices impact the degree to which reported incidence rates are representative in each country. The confidence in reported incidence rates can be used to manage risk through our model, for instance by assigning a source country a confidence range of incidence rates rather than a single value. This helps drive decisions that avoid crossing from one region to another in the contour plot for countries with high uncertainty in their reporting data. |
Processed_Should_international_borders_re-open?_The_impact_o.txt | Another potential source of uncertainty is in estimating incoming travel volumes. Australia’s geography as an island nation adds confidence that the data used to estimate incoming travel volumes is representative of actual travel volume, which in turn increases confidence in the estimated importations in our model. Applying our model to other countries, particularly countries with more porous land borders, needs to consider the greater uncertainty in incoming travel volumes from neighbouring countries. |
Processed_Should_international_borders_re-open?_The_impact_o.txt | The spatial heterogeneity of COVID-19 in source countries is another important factor in our model. For instance, larger countries such as the US or China have experienced surges of COVID-19 in specific states or regions over time. Currently, our model accounts for national incidence rates that cover the entire country. While region-specific incidence rates are available for some countries, incoming travel volumes are reported at country level, which necessitates the averaging of incidence rates across a country. Region-specific travel volumes could add to the granularity of our model. |
Processed_Should_international_borders_re-open?_The_impact_o.txt | Estimating traveller volumes We estimate the expected number of arrivals into Australia, assuming no travel restrictions were implemented, based on five years of historical data (January 2015 - December 2019). To give the data a normal shape, a Box-Cox transformation is applied to each individual time-series19. We use a seasonal autoregressive integrated moving average (SARIMA) model that is suitable for forecasting time-series with seasonal variations that are possibly non-stationary4. To find the model with the best fit we perform a step-wise search over the model space and chose the model with the lowest Akaike Information Criterion20. All calculations have been carried out with Python’s pmdarima statistical library. |
Processed_Should_international_borders_re-open?_The_impact_o.txt | Estimating ascertainment of COVID-19 cases We use the monthly number of observed COVID-19 infections amongst travellers arriving into Australia from a given country to estimate the country’s true incidence rate. We account for Australia’s ascertainment in a similar way by estimating the true incidence rate with the observed incidence in Australian travellers arriving into New Zealand. |
Processed_Should_international_borders_re-open?_The_impact_o.txt | Let γ be the true incidence rate of COVID-19 in a given country. Then, in a sufficiently large sample of the population, we expect to find nγ infected individuals, where n is the size of the sample, i.e. the arrivals from the given source country. Assuming that the number of infected individuals follows a Binomial distribution with unknown parameter γ, the maximum-likelihood estimate of γ is given by ˆγ = x/n, where x is the number of infected individuals in the sample population. We construct the 95% Agresti-Coull interval21 to ensure that the interval falls within the parameter space. The estimated incidence rates and their 95% confidence intervals are shown and compared to those reported by the ECDC in Supplementary Data File 2. |
Processed_Should_international_borders_re-open?_The_impact_o.txt | In addition to adjusting the ECDC data to account for under-reporting, we estimate the disease onset date of all reported cases. From the data published in a recent study22, we find that the delay between a case showing symptoms and being reported follows a Gamma distribution. To adjust the data we subtract X days from the date of report for each recorded COVID-19 case, where X ∼ Gamma(1.76, 4.47). |
Processed_Should_international_borders_re-open?_The_impact_o.txt | where dc denotes the date n − 1 days prior to the arrival date and n is the sum of the latent and the infectious period. |
Processed_Should_international_borders_re-open?_The_impact_o.txt | where dl denotes the date l days prior to the arrival date and l is the length of the latent period. |
Processed_Should_international_borders_re-open?_The_impact_o.txt | Following recent studies of the infectiousness profile of COVID-19, we set the infectious period to eleven days, beginning three days prior to the onset of symptoms23, 24. The incubation period describes the time between contracting a disease and showing symptoms. There is general agreement that the incubation period of COVID-19 is between five and six days25–27 and is approximately three days longer than the latent period23. We draw the length of the incubation period from a log-normal distribution with mean equal to 1.621 and standard deviation equal to 0.41826. To find the latent period we subtract three from the incubation period. |
Processed_Should_international_borders_re-open?_The_impact_o.txt | The number of individuals that an infectious traveller infects while on a plane is drawn from a Poisson distribution with mean λ = tR0/s28, where R0 denotes the basic reproduction rate, s is the length of the infectious period and t is the duration of the flight. We set R0 = 14.8, following the results presented in a study that estimates the basic reproduction rate of COVID-19 on a cruise ship29. We assume R0 on a ship to be similar on a plane where the population is almost fully mixed. |
Processed_Should_international_borders_re-open?_The_impact_o.txt | where X ∼ Poisson(λ ), pi is the probability that individual i is infected and ri is the probability that individual i is infectious during the flight. The sums run over all individuals who arrive during the period of interest. |
Processed_Should_international_borders_re-open?_The_impact_o.txt | COVID-19 importation data for NSW was obtained from NSW Health and is not publicly available. |
Processed_Should_international_borders_re-open?_The_impact_o.txt | We would like to thank Frank de Hoog and Peter Caley for their comments, which greatly helped to improve the manuscript. This work is part of the DiNeMo project. |
Processed_Should_international_borders_re-open?_The_impact_o.txt | JL and KN conceived the study and developed the model. JL performed the analysis. KN, RJ and DP assisted with the analysis and contributed to the interpretation of the results. AES wrote the code to clean and process the data. All authors edited and approved the final manuscript. |
Processed_Water_maser_detections_in_southern_candidates_to_p.txt | Aims. We intended to study the incidence and characteristics of water masers in the envelopes of stars in the post-AGB and PN evo- lutionary stages. Methods. We have used the 64-m antenna in Parkes (Australia) to search for water maser emission at 22 GHz, towards a sample of 74 sources with IRAS colours characteristic of post-AGB stars and PNe, at declination < −32o. In our sample, 39% of the sources are PNe or PNe candidates, and 50% are post-AGB stars or post-AGB candidates. Results. We have detected four new water masers, all of them in optically obscured sources: three in PNe candidates (IRAS 12405−6219, IRAS 15103−5754, and IRAS 16333−4807) and one in a post-AGB candidate (IRAS 13500−6106). The PN can- didate IRAS 15103−5754 has water fountain characteristics, and it could be the first PN of this class found. Conclusions. We confirm the tendency suggested in Paper I that the presence of water masers during the post-AGB phase is favoured in obscured sources with massive envelopes, in comparison with objects with optical counterparts in these phases. We propose an evolutionary scenario for water masers in the post-AGB and PNe stages, in which “water fountain” masers could develop during post-AGB and early PN stages. Later PNe would show lower velocity maser emission, both along jets and close to the central objects, with only the central masers remaining in more evolved PNe. |
Processed_Water_maser_detections_in_southern_candidates_to_p.txt | The study of H2O masers in evolved stars has been proved to be a very powerful tool to understand the evolution and shap- ing of sources between the Asymptotic Giant Branch (AGB) and the phase of Planetary Nebula (PN). About 80% of the stars (those with M≤8M⊙) go through these evolutionary stages, where a dramatic change in their morphology is produced. The stars in the AGB stage have a high mass-loss rate, that can reach ≃ 10−4 M⊙yr−1. This mass-loss is usually spherical, although the first manifestations of asymmetry are already observed in some of them (see for example Le˜ao et al. 2006; Josselin et al. 2000). A large fraction of PNe show, however, axial symmetries with bipolar or multipolar morphologies and, in many cases, jets (Sahai et al. 2007). |
Processed_Water_maser_detections_in_southern_candidates_to_p.txt | are located at the inner part of the envelope (between 10 and 100 AU) and OH masers at distances > 103 AU from the central star (Reid & Moran 1981). |
Processed_Water_maser_detections_in_southern_candidates_to_p.txt | In the classical view of maser emission pumped by spherical AGB winds, SiO masers are expected to survive only ≃ 10 yr after the stop of the strong mass-loss, while H2O and OH masers will be extinguished after ≃ 100 yr and ≃ 1000 yr, respectively (Lewis 1989; G´omez et al. 1990). These timescales imply that in the post-AGB stage we can still find some objects that harbour water masers with the same characteristics of those found in the AGB stage (see for example Engels 2002). On the other hand, water maser emission was not expected in PNe. |
Processed_Water_maser_detections_in_southern_candidates_to_p.txt | gesting that the asymmetrical mass loss can be related to the presence of water masers in PNe. |
Processed_Water_maser_detections_in_southern_candidates_to_p.txt | In order to study the possible relationships between the evo- lutionary stage of evolved objects and the presence of water masers, Su´arez et al. (2007) (hereafter Paper I) carried out a single-dish survey for H2O maser emission towards the evolved sources in the catalogue of Su´arez et al. (2006) (hereafter SGM) that were located north of declination −32o. A total of 105 sources were searched with the Robledo 70m antenna, with five detections, four of them reported in that paper for the first time. Three of these detections were found in obscured sources, of which one was a water fountain (confirmed by Su´arez et al. 2008) and another one a PN candidate. These results led to sug- gest that the presence of water masers in the post-AGB and PN phases is favoured in stars with massive envelopes, and that PNe harbouring water masers are a special class of massive, rapidly evolving PNe. Water masers in these sources would be related to processes of non-spherical mass-loss, rather than being the rem- nant of water masers pumped by spherical winds in the AGB stage. In Paper I it was also suggested that probably all H2O- PNe had traversed previously by a water fountain phase (see also G´omez et al. 2008). The discovery of objects showing both water fountain and PNe characteristics would be an important evidence of this evolutionary path. |
Processed_Water_maser_detections_in_southern_candidates_to_p.txt | In this paper we present the continuation of the work started in Paper I, with single-dish observations of sources in the SGM catalogue with declination ≤ −32o, using the Parkes antenna. This paper is structured as follows: In Sec. 2 we present the sam- ple of observed sources. In Sec. 3 we describe the observations performed. In Sec. 4 we describe the results obtained, including a description of the detected objects. In Sec. 5 we discuss our results and finally, in Sec. 6 we give the main conclusions of this work. |
Processed_Water_maser_detections_in_southern_candidates_to_p.txt | The selection criteria are described in detail in Paper I and in SGM. Briefly, the SGM atlas is composed of IRAS sources lo- cated in the region of the IRAS colour-colour diagram popu- lated mainly by post-AGB stars (van der Veen et al. 1989). The sources in this region would have expanding envelopes charac- terised by a range of dust temperatures between 200 K and 80 K, and radii between 0.01 pc and 0.1 pc. |
Processed_Water_maser_detections_in_southern_candidates_to_p.txt | The evolved sources in the SGM catalogue are classified in five different groups: PN, post-AGB stars, transition sources (that are evolving between the post-AGB and PN phases), pecu- liar sources (evolved sources with peculiar characteristics), and objects with no optical counterpart (which could be evolved ob- jects in any of the stages mentioned above, but could no be clas- sified with the optical spectroscopy presented in SGM). |
Processed_Water_maser_detections_in_southern_candidates_to_p.txt | For these objects with no optical counterpart found by SGM, we can only give a tentative classification as “post-AGB can- didate” or “PN candidate”. We consider them candidates to PN when there is an indication of the ionisation of the envelope, such as the presence of radio continum emission. In many cases, these PN candidates are merely point-like sources in infrared images, with no further evidence of the presence of a nebula. When there is no indication of ionisation or nebulosity, we consider these obscured objects as post-AGB candidates. Obviously, further in- frared observations are required to ascertain the true nature of these objects without optical counterpart. |
Processed_Water_maser_detections_in_southern_candidates_to_p.txt | By the time the observations described in Sec. 3 were per- formed, we became aware of the paper by Deacon et al. (2007), who carried out a survey for water masers in post-AGB can- didates with OH 1662 MHz maser emission. Their survey in- cluded 7 of the sources selected for our observations: IRAS 14341−6211, IRAS 17088−4221, IRAS 17164−3226, IRAS 17168−3736, IRAS 17245−3951, IRAS 17310−3432, IRAS 17370−3357, with one water maser detection among them (IRAS 17088−4221). Therefore, these 7 sources were explicitly left out of our observations, although we will include them in the discussion of this paper (see Sec. 5), whenever appropriate, because they fulfil the criteria used to select the sample. |
Processed_Water_maser_detections_in_southern_candidates_to_p.txt | We observed the 616 → 523 transition of the water molecule (rest frequency 22235.080 MHz), using the 64m antenna of the Parkes Observatory1, on 2007 February 08-12. At this frequency, the half-power beamwidth of the telescope is ≃ 1.′3. The 1.3 cm receiver of this antenna comprises a cryogenic High Mobility Electron Transistor. Both right and left circular polarisation were observed simultaneously. As a backend, we used the Multibeam Correlator, covering a bandwidth of 64 MHz (i.e., velocity cov- erage ≃ 862.9 km s−1) with 2048 spectral channels for each polarisation, thus yielding a spectral resolution of 31.25 kHz (0.42 km s−1). Since little information about the radial velocity of these sources is available, all observed spectra were centred at vLSR = 0 km s−1. The observations were taken in position- switching mode with a total integration time (on+off) of 30 min. The rms pointing accuracy of the telescope was ≃ 10′′. |
Processed_Water_maser_detections_in_southern_candidates_to_p.txt | the 7 objects observed by Deacon et al. (2007) (1 PN and 4 post-AGB stars with optical counterpart, and 2 post-AGB candidates without optical counter- part) comprised all objects without optical counterpart as well as all PNe and transition objects with optical counterpart included in the selected sample of 106 sources from SGM. |
Processed_Water_maser_detections_in_southern_candidates_to_p.txt | Fig. 1. Water maser spectra of the detected sources. The numbers shown at the upper left of the graph correspond to the day of observation, in 2007 February. |
Processed_Water_maser_detections_in_southern_candidates_to_p.txt | observations, all of them reported here for the first time: IRAS 12405−6219, IRAS 13500−6106, IRAS 15103−5754, and IRAS 16333−4807. Their spectra are shown in Figs. 1 and 2. All of them correspond to objects classified as “sources without opti- cal counterpart” by SGM. |
Processed_Water_maser_detections_in_southern_candidates_to_p.txt | IRAS 15103−5754, and by classified van de Steene & Pottasch (1993) as PN candidates due to the presence of radio continuum emission, while IRAS 13500−6106 has been classified as a post-AGB star by Garc´ıa-Lario et al. (1997), based on its position in the IRAS two-colour diagram and on its low IRAS variability index (v=29). |
Processed_Water_maser_detections_in_southern_candidates_to_p.txt | In the following we will describe in detail the four detected objects and comment on their nature. In order to identify pos- sible candidates to the maser-exciting source, we have also in- spected MSX and 2MASS data. |
Processed_Water_maser_detections_in_southern_candidates_to_p.txt | by at van de Steene & Pottasch Dec.(J2000)=−58o05′21.1′′, R.A.(J2000)=15h14m18.53s, i.e., 15′′ from the IRAS nominal position, with a flux density S (6cm) = 115 mJy. OH maser emission at 1612 MHz has been detected in this source (Te Lintel Hekkert & Chapman 1996) at VLSR ≃ −47 km s−1 with a flux density ∼0.3 Jy. The only MSX source in the beam of our observations is MSX6C G320.9062-00.2928, that is coincident with the 2MASS source J15141845-5805203 (within ∼ 1.′′2). The MSX source is also coincident with the radio continuum position (within ∼ 1.′′4). These results suggest that IRAS 15103−5754 as a very likely PN, associated with the mentioned IR sources. |
Processed_Water_maser_detections_in_southern_candidates_to_p.txt | Fig. 2. Close up view of the water maser spectrum taken on 2007 February 9 toward the PN candidate IRAS 15103-5754. |
Processed_Water_maser_detections_in_southern_candidates_to_p.txt | We detected water maser emission with a single spectral com- ponent at VLSR ≃ −23.6 km s−1 and a flux density of ∼1 Jy (Fig. 1 and Table 2). A source of continuum emission at 6 cm was detected by van de Steene & Pottasch (1993) at R.A.(J2000)=12h43m32.04s, Dec.(J2000)=−62o36′13.9′′, with a flux density S (6cm) = 14 mJy. This radio source is located 6′′ from the IRAS nominal position. |
Processed_Water_maser_detections_in_southern_candidates_to_p.txt | The high velocity range found in the water maser compo- nents makes this source a new candidate to the water fountain class. Moreover, if its classification as a PN and its association with the water maser emission are confirmed, this source would be the first PN showing water fountain characteristics. The ex- istence of “water fountain-PNe” was suggested by G´omez et al. (2008). They discussed the possibility of finding a PN that would still maintain the high-velocity water masers present during its previous post-AGB−water fountain stage. These objects should be among the youngest and more massive PNe known, and they would be excellent targets to study the “real-time” evolution of a PN, since the time scales of evolution of this type of sources would be very short and can be studied by means of high angular resolution observations of the water masers. |
Processed_Water_maser_detections_in_southern_candidates_to_p.txt | There is only one MSX source in the beam of the water maser observations. This source (MSX6C G302.0213+00.2542) is located ∼ 6′′ from the reported radio continuum position, and is coincident with the 2MASS source J12433151-6236135. We can consider that the radio continuum is associated with the IR source within the errors. The water maser position is also prob- ably associated with this source. |
Processed_Water_maser_detections_in_southern_candidates_to_p.txt | Although van de Steene & Pottasch (1993) classify it as a possible PN, this source also fulfils the criteria used by Hughes & MacLeod (1989) to select HII regions from the IRAS ≥ 0.25 and Point Source Catalogue (log S60µm/log S25µm log S25µm/log S12µm ≥ 0.4). We will thus maintain its classifi- cation as PN as doubtful. |
Processed_Water_maser_detections_in_southern_candidates_to_p.txt | We have detected water maser emission towards this source (see Fig. 1 and Table 2) with a maximum flux density of ∼1.9 Jy at VLSR ≃ 45 km s−1. The radio continuum source detected by van de Steene & Pottasch (1993) is located at R.A.(J2000)=16h37m06.79s, Dec.(J2000)=−48o13′42.6′′, dis- placed ≃ 7′′ from the IRAS nominal position, with a flux density S (6cm) = 32.5 mJy. |
Processed_Water_maser_detections_in_southern_candidates_to_p.txt | The only MSX source in the beam of our observations is MSX6C G336.6445-00.6953, that is coincident with the source found by van de Steene & Pottasch (1993) (within ∼ 1.′′4) and with the 2MASS source J16370659-4813429 (within ∼ 1.′′5). The likely association of the radio continuum and IR sources makes this object a plausible PN candidate. |
Processed_Water_maser_detections_in_southern_candidates_to_p.txt | Our water maser spectrum shows one component with a max- imum flux density of ∼ 1.5 Jy at VLSR ≃ 40.0 km s−1 (see Fig. 1 and Table 2). This source was classified as a post-AGB star by Garc´ıa-Lario et al. (1997) based on its IR colours. The only MSX source in the beam of the water maser observations is MSX6C G310.3099+00.6306, and there is no 2MASS source associated with it. No OH has been detected in this source by Te Lintel Hekkert et al. (1991). |
Processed_Water_maser_detections_in_southern_candidates_to_p.txt | We note that IRAS 17580−3111 and IRAS 17443−2949 were initially classified as obscured PNe with OH emission (OHPN) by Zijlstra et al. (1989), and as such were cited in Su´arez et al. (2007), but G´omez et al. (2008) showed that the radio continuum reported by Ratag et al. (1990) was not asso- ciated with these sources, and Garc´ıa-Hern´andez et al. (2007) classified them as post-AGB and AGB stars respectively. We will adopt the classification of Garc´ıa-Hern´andez et al. (2007) for the discussion. |
Processed_Water_maser_detections_in_southern_candidates_to_p.txt | In summary, out of the nine detected sources, only one turned out to be a “classical” post-AGB star with an optical counterpart. The rest of the possible post-AGB stars in wich H2O emission was found do not have a counterpart in the visible domain. The fact that these stars could already be in the post-AGB stage while still having a thick envelope that prevents their detection in the optical, suggests that they are relatively massive stars that have quickly evolved from the AGB. The new detections found in this campaign confirm the trend already found in Paper I, relating the presence of water masers in post-AGB stars or PN to optically obscured sources. |
Processed_Water_maser_detections_in_southern_candidates_to_p.txt | Apparently, the incidence of water maser emission during the post-AGB phase is higher in obscured objects, i.e. those with more massive envelopes. Of the 80 visible post-AGB stars ob- served, only 1 turned out to have water maser emission (∼1%), while we have found this emission in 4 out of 35 post-AGB star candidates, without optical counterpart (∼11%). This suggests that the presence of massive envelopes favour the pumping of water masers during the post-AGB stage. Morever, our water maser detections towards three obscured PN candidates might suggest a similar trend during the PN phase. The detection of dense molecular gas toward an H2O-PN (Tafoya et al. 2007) also supports the existence of massive envelopes in these objects. |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.