COVID-19 short-term forecasts Deaths 2020-04-17


Disclaimer

  • Forecasts produced by Jennie Castle, Jurgen Doornik, and David Hendry, researchers at the University of Oxford. These are our forecasts, and should not be considered official forecasts from, or endorsed by, any of: University of Oxford, Oxford Martin School, Nuffield College, or Magdalen College.
  • These forecasts are short term time-series extrapolations of the data. They are not based on epidemiological modelling or simulations. The documentation that is provided is still in progress and not peer reviewed. All forecasts are uncertain: their success can only be determined afterwards. Many mitigation strategies are in place, which, if successful, invalidate our forecasts. An explanation of our methods is provided below.

Recent changes

[2020-03-24] Our forecasts are starting to overestimate in some cases. This was always expected to happen when the increase starts to slow down. Scenario forecasts that are based on what happened in China earlier this year, but only for Italy and Spain sofar.
[2020-03-26] Scenario forecasts that are based on what happened in China earlier this year, only for Italy.
[2020-03-31] Scenario forecasts, based on what happened in China earlier this year, are presented for several countries (line marked with x). Created more plausible 90% confidence bands (dotted line in same colour).
[2020-04-02] Now including more US States, based on New York Times data. And the world.
[2020-04-06] Added a post hoc estimate of the peak number of cases. This needs at least three confirmed observations (four for deaths) after the event. It is based on the averaged smooth trend, and can change later or be a local peak. It is marked with a vertical line with the date label, or a date with left arrow in the bottom left corner of the graph. This is backported to 2020-04-04.
[2020-04-08] Minor correction to peak estimates. Added table with scenario forecasts.
[2020-04-09] Added table with estimated peak dates (if happened) and dates to and since the peak. Note that this can be a local peak, and subsequent re-acceleration (or data revisions) can result in a new peak later.
[2020-04-10] Updated documentation with better description of short-term estimates and peak determination.
[2020-04-16] Added scenario forecasts to all graphs now. This would now be the preferred forecast for most.
This is the first time with a peak in confirmed UK cases (also for deaths, but this is uncertain because it is at the same date).
[2020-04-17] Bird and Nielsen look into nowcasting death counts in England.

Further information

  • We believe these forecasts fill a useful gap in the short run. They give an indication of what is likely to happen in the next few days, removing some aspect of surprise. Moreover, a noticeable drop in comparison to the extrapolations could be an indication that the implemented policies are having some impact. It is difficult to understand exponential growth. We hope that these forecasts may help to convince viewers to adhere to the policies implemented by their respective governments, and keep all arguments factual and measured.
  • We use the data repository for the 2019 Novel Coronavirus Visual Dashboard operated by the Johns Hopkins University Center for Systems Science and Engineering. This is updated daily, but we tend to update our forecasts only every other day.
    US state data as of 2020-03-28 is courtesy of the New York Times.
  • We can only provide forecasts of what is measured. If confirmed cases are an underestimate of actual cases, then our forecasts will also be underestimates. No other epidemiological data is used. Data definition and collection differs between countries and may change over time.
  • We will update the methodology as we learn what is happening in the next few days or weeks. Once the number of cases levels off, there is no need to provide these forecasts anymore.
  • Countries where the counts are very low or stable have been omitted.
  • The graphs have dates on the horizontal axis (yyyy-mm-dd) and cumulative counts on the vertical axis. They show
    1. bold dark grey line (with circles): observed counts (Johns Hopkins CSSE);
    2. many light grey lines (with open circles): forecasts using different model settings and starting up to four periods back;
    3. red line (with open circles): single forecasts path using default model settings;
    4. black line (with crosses): average of all forecasts, recentered on the last observation;
    5. thin green lines: some indication of uncertainty around the red forecasts, but we do not know how reliable that is.
    Both the red line forecasts and the black lines are also given in the tables above. These forecasts differ, we are currently inclined to use the average forecasts.
  • The forecasts are constructed as follows:
    1. An overall `trend' is extracted by taking a window of the data at a time. In each window we draw `straight lines' which are selected using an automatic econometric procedure (`machine learning'). All straight lines are collected and averaged, giving the trend.
    2. Forecasts are made using the estimated trend, but we note that this must be done carefully, because simply extrapolating the flexible insample trend would lead to wildly fluctuating forecast. We use the `Cardt' method, which has been found to work well in other settings.
    3. Residuals from the trend are also forecast, and combined with trend forecasts into an overall forecast.
  • Scenario forecasts are constructed very differently: smooth versions of the Chinese experience are matched at different lag lengths with the path of each country. This probably works best from the peak, or the slowdown just before (but we include it for the UK nonetheless).
  • The forecast evaluation shows past forecasts, together with the outcomes (in the grey line with circles).
  • EU-BS is Estonia, Latvia, and Lithuania together.
  • This paper describes the methodology and gives further references, but remains preliminary. Also preliminary is the documentation of the medium term forecasts.

Deaths count average forecast Europe (bold black line in graphs) 2020-04-18 to 2020-04-24

DateUKEUATBEDEDKESFRIEITNLPLPTROSECH
2020-04-17 14576 79333 431 5163 4352 336 20002 18681 530 22745 3459 332 657 411 1400 1327
2020-04-18 15400 82000 450 5480 4660 350 20500 19500 570 23300 3610 350 690 430 1510 1370
2020-04-19 16200 85000 460 5780 4950 360 21000 20300 610 23800 3730 370 720 450 1630 1410
2020-04-20 17000 88000 470 6100 5260 370 21500 21100 650 24300 3860 390 760 470 1770 1460
2020-04-21 17800 91000 490 6430 5590 380 21900 22000 700 24900 3990 420 790 480 1910 1500
2020-04-22 18700 94000 500 6780 5940 400 22400 22900 740 25400 4120 440 830 500 2060 1550
2020-04-23 19600 97000 520 7160 6310 410 22900 23800 800 26000 4260 460 870 520 2230 1600
2020-04-24 20600 100000 530 7560 6710 420 23400 24800 850 26600 4410 490 910 550 2410 1650

Deaths count average forecast (bold black line in graphs) 2020-04-18 to 2020-04-24

DateBrazilCanadaIranPhilippinesUSUS-CAUS-COUS-CTUS-FLUS-GAUS-ILUS-LAUS-MAUS-MIUS-NJUS-NYUS-WA
2020-04-17 2141 1354 4958 387 36773 1050 389 1036 726 661 1142 1213 1404 2226 3840 12822 610
2020-04-18 2360 1510 5050 410 40700 1140 410 1150 780 710 1230 1290 1560 2380 4220 13500 630
2020-04-19 2600 1650 5140 430 44400 1230 440 1260 830 760 1320 1380 1720 2530 4640 14200 650
2020-04-20 2860 1810 5240 450 48500 1320 470 1380 890 810 1420 1480 1890 2690 5110 14800 660
2020-04-21 3140 1980 5340 480 53000 1420 500 1510 950 870 1520 1580 2080 2870 5620 15600 680
2020-04-22 3460 2170 5440 510 57900 1530 530 1660 1020 930 1630 1690 2290 3050 6180 16300 700
2020-04-23 3790 2370 5540 540 63300 1640 560 1810 1090 990 1750 1810 2530 3250 6790 17100 720
2020-04-24 4170 2600 5650 570 69300 1770 600 1990 1170 1060 1880 1930 2780 3460 7470 18000 740

Deaths count forecast Europe (bold red line in graphs) 2020-04-18 to 2020-04-24

DateUKEUATBEDEDKESFRIEITNLPLPTROSECH
2020-04-17 14576 79333 431 5163 4352 336 20002 18681 530 22745 3459 332 657 411 1400 1327
2020-04-18 15300 82000 440 5390 4650 350 20400 19400 570 23200 3590 350 690 430 1510 1360
2020-04-19 16000 85000 460 5660 4990 360 20900 20100 610 23700 3730 370 710 440 1630 1400
2020-04-20 16700 88000 470 5940 5370 370 21400 20900 660 24200 3870 390 740 460 1760 1440
2020-04-21 17400 91000 490 6220 5770 380 21900 21800 700 24800 4010 410 770 480 1890 1490
2020-04-22 18100 94000 500 6510 6200 390 22400 22600 750 25300 4160 430 810 490 2040 1530
2020-04-23 18800 97000 520 6810 6670 410 23000 23500 810 25900 4310 450 840 510 2200 1580
2020-04-24 19600 100000 530 7120 7170 420 23500 24400 870 26400 4470 470 870 530 2360 1620

Deaths count forecast (bold red line in graphs) 2020-04-18 to 2020-04-24

DateBrazilCanadaIranPhilippinesUSUS-CAUS-COUS-CTUS-FLUS-GAUS-ILUS-LAUS-MAUS-MIUS-NJUS-NYUS-WA
2020-04-17 2141 1354 4958 387 36773 1050 389 1036 726 661 1142 1213 1404 2226 3840 12822 610
2020-04-18 2450 1510 5050 400 40300 1160 450 1160 770 690 1190 1290 1640 2360 4220 13400 630
2020-04-19 2800 1670 5130 420 44200 1270 500 1270 830 730 1240 1370 1910 2510 4650 14000 650
2020-04-20 3190 1840 5220 440 48500 1390 550 1390 890 770 1290 1450 2220 2660 5120 14600 660
2020-04-21 3620 2020 5310 460 53200 1520 600 1520 950 820 1340 1540 2570 2830 5630 15200 680
2020-04-22 4110 2230 5400 470 58300 1660 670 1660 1020 860 1400 1640 2960 3000 6180 15800 700
2020-04-23 4660 2450 5490 490 64000 1810 730 1810 1090 910 1460 1740 3420 3190 6790 16400 720
2020-04-24 5290 2700 5580 510 70200 1980 810 1980 1170 960 1520 1850 3940 3390 7460 17000 740

Deaths count scenario forecast (bold green line in graphs) 2020-04-18 to 2020-04-26

DateUKEUATBEDEDKESFRIEITNLPLPTROSECHBrazilCanadaIranPhilippinesUSUS-CAUS-COUS-CTUS-FLUS-GAUS-ILUS-LAUS-MAUS-MIUS-NJUS-NYUS-WA
2020-04-17 14576 79333 431 5163 4352 336 20002 18681 530 22745 3459 332 657 411 1400 1327 2141 1354 4958 387 36773 1050 389 1036 726 661 1142 1213 1404 2226 3840 12822 610
2020-04-18 15200 81200 440 5370 4540 340 20400 19300 560 23200 3560 350 680 430 1500 1360 2290 1450 5040 400 40200 1140 400 1120 770 680 1220 1290 1540 2340 4110 13500 620
2020-04-19 15900 83500 450 5630 4760 350 20700 19900 600 23700 3670 370 710 440 1610 1400 2400 1540 5110 420 42300 1220 420 1140 820 710 1290 1370 1660 2450 4380 14100 630
2020-04-20 16500 85800 470 5910 4990 360 21100 20500 630 24100 3760 380 730 460 1700 1440 2540 1630 5180 430 43500 1300 430 1140 860 730 1360 1440 1790 2570 4650 14700 650
2020-04-21 17100 88000 480 6090 5210 360 21400 21200 660 24500 3840 400 760 480 1760 1470 2680 1720 5240 450 46000 1380 450 1140 910 740 1420 1510 1920 2670 4910 15200 660
2020-04-22 17700 90100 490 6290 5430 370 21600 21800 700 24900 3920 410 780 490 1830 1500 2820 1800 5300 460 47700 1470 460 1140 950 760 1490 1580 2050 2780 5170 15800 660
2020-04-23 18300 92200 510 6610 5640 370 21900 22400 730 25300 4000 430 800 500 1870 1530 2960 1880 5360 480 49800 1540 480 1140 1000 780 1560 1650 2180 2890 5420 16300 670
2020-04-24 18800 94100 520 6850 5830 380 22200 22900 760 25700 4090 440 820 520 1940 1550 3090 1960 5420 490 52200 1620 490 1140 1040 800 1640 1710 2310 3000 5660 16900 680
2020-04-25 19400 96000 520 7070 6030 380 22400 23400 790 26100 4170 460 850 530 2010 1570 3220 2040 5480 510 54000 1690 510 1140 1080 820 1720 1780 2430 3100 5900 17400 690
2020-04-26 19900 97700 530 7250 6220 390 22700 23800 810 26500 4230 470 870 540 2080 1590 3340 2110 5530 520 56100 1760 520 1140 1120 850 1770 1840 2550 3180 6120 17800 690

Peak increase in estimated trend of Deaths in Europe 2020-04-17

UKEUATBEDEDKESFRIEITNLPLPTROSECH
Peak date04-1004-0704-0804-10 --04-0404-0204-07 --03-2804-0704-1204-1104-11 --04-04
Peak daily increment 883 3208 23 361 19 907 1021 826 159 21 33 24 61
Days from 100 to peak 22 34 9 17 3 20 23 24 18 6 13 9 12
Days from peak/2 to peak 16 24 17 15 14 16 17 19 18 18 19 17 17
Days since peak 7 10 9 7 13 15 10 20 10 5 6 6 13

Peak increase in estimated trend of Deaths 2020-04-17

BrazilCanadaIranPhilippinesUSUS-CAUS-COUS-CTUS-FLUS-GAUS-ILUS-LAUS-MAUS-MIUS-NJUS-NYUS-WA
Peak date -- --03-1904-13 -- -- -- -- -- -- -- -- --04-11 --04-1004-06
Peak daily increment 150 28 142 786 30
Days from 100 to peak 14 12 14 19 15
Days from peak/2 to peak 16 26 15 15 30
Days since peak 29 4 6 7 11

Initial visual evaluation of forecasts of Deaths