County Pane
Selecting a county displays the "Cumulative Cases" graph to show the past seven days of confirmed COVID-19 cases based on our data source, and if available, IEM’s projected estimates for the next seven days. The orange "whisker" vertical lines on each forecast indicate the high and low projections. The centerline provides the most probable values.
State Heatmap
Selecting a county displays the "Cumulative Cases" graph to show the past seven days of confirmed COVID-19 cases based on our data source. If available, IEM's projected estimates for the next seven days are included. The orange "whisker" vertical lines on each forecast indicate the high and low projections. The centerline provides the most probable values.
While IEM does calculate projections for all 50 states, Puerto Rico, the U.S. Virgin Islands, Guam, and select counties, not all jurisdictions throughout the United States and its territories currently have forecasted data.
National Heatmap
The national map displays historical data (blue) and forecast data (orange/red) at the county level. This data is available as either confirmed or per capita cases, which is selectable from the drop-down above the national map.
By clicking a county on the map, the state and county panes will update. Moving the slider bar allows viewing the data set over time. Click the arrows on the slider to change the date by a single day.
By default, when the site loads, the national and state maps display data from the last forecast date available. Today’s value is also a forecast number as the reported cases for today are not yet known. IEM is confident in its AI modeling. Presenting the forecast data first allows us to stand apart from other sites.
Depending on when the data is made available and on the speed of the processing/review cycles, the "Updated" date at the top of the page may not always match the current date. This update methodology may also cause the data to be more than a day old (in some cases, as many as three days old).
Top 20 List
Displays the top 20 counties that have the highest number of new COVID-19 cases from the previous day. The list defaults to the national rankings, but once a state is selected, the top 20 shows the county rankings for that state. The three-day average of new cases is in parentheses. If the previous day’s value is less than the three-day average, the number will be green. If not, the number will be red.
Data Overview
The IEM COVID-19 projections use an artificial intelligence (AI) based disease model to best predict the likelihood of new COVID-19 cases during the next seven days. The resulting seven day projections are typically within 10%—often within 5%—of actual confirmed cases, in part because a very large number of simulations is used to help account for data shortfalls, e.g., delays in reporting, under-reporting. Also, this seven-day window provides reasonable accuracy before the degree of uncertainty grows too large.
IEM’s AI model uses a hurricane modeling approach that works "backward" to find the best modeling parameters based on actual data in order to develop future projections. Rather than selecting modeling parameters to try and match the confirmed data, the data is analyzed "backwards" using over 11 million simulations to calculate the set of modeling parameters that best fit what has happened in the real world. These parameters are then used in thousands of "forward" modeling simulations to project the number of new COVID-19 cases over the next seven days.
Our data source has recently undergone significant changes in county level COVID-19 case reporting in Massachusetts and Utah. For several counties in those two states reported case data is not available. Consequently only state-level and available county-level data and projections are available on this site.
