How do forecasters predict the weather? They look at the weather happening now and determine how it will change over time. Comprehensive and current weather data means more accurate predictions, and data that updates in real time is best.In this lesson, you'll create a map of real-time weather data.
![](/uploads/1/2/6/3/126329206/193113772.jpg)
You'll investigate trends, learn about atmospheric processes, and predict weather. For areas without weather stations, you'll interpolate data to make decisions.This lesson is an introduction to real-time data and major weather concepts, targeted toward students. The map navigates to the extent you were at when you created the bookmark.Add dataNext, you'll add data to the map. Your data will come from several sources, including NOAA and the ArcGIS Living Atlas of the World, Esri's curated collection of geographic information from around the world. These sources are authoritative, so you know the data will be accurate.First, you'll add another basemap, the World Light Gray Base.
This basemap contains less geographic information than the Topographic basemap, so it'll emphasize the real-time weather data. However, you want the ability to turn it off if more geographic context is needed, so you'll add it as a layer, not a basemap.
The layer's metadata opens. It contains detailed descriptions of the agency responsible for the satellite, the sensors that produce the imagery, and how often it is updated (every 30 minutes).The imagery not only shows what is visible to the human eye, but also infrared light. Infrared sensors help show the relative warmth of objects, which is important for determining the temperature of clouds. In this imagery, white clouds are colder, while dark clouds are warmer. Close the metadata.
I would do this in two stages. Create a raster by interpolating your heights (see here). Ensure the raster resolution is the same as the point.
In the Contents pane, for sat meteo imagery time, click More Options and choose Rename. Rename the layer GOES Satellite Imagery and click OK. Metadata for this layer explains that this layer was created by combining reflectivity radar data from next-generation radar (NEXRAD) locations across the United States.These radars work by sending out radar waves that reflect off precipitation. Based on how much of the radar's power is reflected, how long the signal took to return to the radar, and how much the radar's frequency changes, the radar can determine the location and intensity of precipitation. Areas that are blue or light green generally indicate light rainfall, while darker greens, yellows, oranges, and reds tend to indicate increasingly more severe rainfall.The data does not extend far from the United States national boundaries and updates every 4 minutes. Rename the layer NEXRAD Precipitation. Uncheck the layer to turn it off.
Note:Depending on the time of year when you take this lesson, there may be no active hurricanes (the Atlantic hurricane season goes from June to November). If no hurricanes are active, you can skip the steps related to hurricane data.The orange points show the observed track of the hurricane. The black points and text are the forecasted track of the hurricane. The gray area indicates the possible margin of error in the forecast. The data was compiled by NOAA but is located on the Living Atlas.
The layer updates every 15 minutes. The legend indicates which arrow colors represent which range of wind speeds. Click the Show Contents of Map button to return to the Contents pane.
Rename the NOAA METAR current wind speed direction layer NOAA METAR Wind Speed and Direction.Create additional layersYou've added a lot of data. However, you don't have layers that represents air pressure and temperature, essential components of weather and weather analysis.
Fortunately, the wind speed and direction layer also contains data on these components. You'll examine the available data, copy the layer, and symbolize the copies to show the appropriate parameter.Although the copies will have the same data as the original, by making three different layers it'll be easier to visualize and analyze the data.
Predict weatherPreviously, you created a map of real-time weather data collected from satellites, radars, and weather stations around the world. You learned a little bit about how the data was collected and who collected it, but in this lesson, you'll go on a deeper exploration of your data and use what you learn to predict weather across time and space. Explore temperatureFirst, you'll look for trends in your data. Accurate weather forecasting is dependent on seeing what patterns are happening now.
You'll start by examining your temperature layer. Tip:To see the temperature ranges each symbol represents, open the Legend pane.In the example image, temperatures seem to be strongly affected by latitude (your temperatures may vary). At least in the United States, the hottest temperatures are in the south, with the coldest temperatures in the north.
However, the correlation between temperature and latitude is not exact, and some states on the same latitude have much different temperatures.You'll explore worldwide to see what other patterns you can find. Navigate to the West Europe bookmark. Which areas have the most data? Which areas have the least?. Name two areas where latitude is the most likely explanation for observed temperature. Name two areas that will likely experience much different temperatures three months from now.
Name one area that will likely experience the same temperatures three months from now. Are there any areas where altitude is causing temperatures that aren't explained by latitude or seasonal variation?. What effect might oceans and large water bodies have on temperature?
How do temperatures tend to differ between coastal and inland areas at the same latitude?. How does the temperature where you live compare to the temperature in the surrounding area? Does the recorded temperature match the temperature you're currently experiencing? If not, what might have caused the difference?Next, you'll calculate statistics to find out the range of temperatures around the world. In the Contents pane, point to the NOAA METAR Temperature Stations sublayer and click Show Table. In the table, click the Air Temperature field and choose Statistics. Note:When you sort the row in ascending order, the first few rows may have no value, meaning the station did not record a temperature the last time the data updated.
You may need to scroll down the table until you find the coldest temperature. Where is this station located?.
Why is this station so cold? Is it primarily cold because of latitude, elevation, or season?.
Close the table.Predict rain with wind and pressureNext, you'll take a closer look at your precipitation and pressure data. In conjunction with your other data layers, you'll predict where rainfall might occur in the near future.First, you'll compare current rainfall to wind patterns to see where the wind might cause rain clouds to travel in the near future. Then, you'll learn how pressure can affect precipitation and determine areas with high and low pressure systems. Does precipitation tend to occur where the clouds are bright white (cold) or dark gray (warm)?In general, warm clouds are lower to the ground and absorb moisture through evaporation. As they rise, they become colder, and the water vapor condenses into liquid droplets. This makes the cloud heavier, so it drops and releases the liquid in the form of precipitation. Turn off the GOES Satellite Imagery layer.
Turn on the NOAA METAR Wind Speed and Direction layer and change the basemap to Topographic. In the example image, southern Louisiana and Mississippi are experiencing a lot of rainfall. The labels for the wind speed features indicate the speed of the wind in kilometers per hour. While not all the wind speed arrows point in the same direction, the overall wind pattern is toward the north and east. If this wind persists, the city of Alexandria might experience rain soon. But how soon?.
Find a city that is currently dry but, based on wind direction, might experience rain soon. On the ribbon, click Measure, choose Distance, and set the units to Kilometers. In the example image, a northeastern arrow with a wind speed of 17 kilometers per hour is about 180 kilometers away from Alexandria. At this rate, it would take over 10 hours for rain to reach the city. Additionally, other stations in the area record either no wind, slower wind, or wind that is more easterly.
It's possible the precipitation will pass south of the city altogether. How far away is rainfall from the city you found?. How long would it take rainfall to reach the city given the wind speed and direction?. Are there other winds that might cause the rainfall to avoid your city?. Overall, how likely would you say it is that your city receives rain?Wind is not the only factor that influences precipitation. Pressure is also important. Low pressure causes air to rise, cool, and condense into rain clouds.
High pressure causes air to flow down and heat up. In the Northern Hemisphere, air tends to move counterclockwise around a low pressure system and clockwise around a high pressure system (this trend is reversed in the Southern Hemisphere).Based on your satellite imagery, precipitation, and wind speed layers, you'll predict where pressure is high and low. Close the Measure window. Navigate to the Continental United States bookmark and turn on the GOES Satellite Imagery Transparent layer.
The map updates with the new symbology. The default color scheme of light blue to dark blue is fine, but the distribution of the data into each symbol class is skewed by the way the data was reported.Air pressure is usually between 1,000 and 1,030 millibars, but stations that did not report any data are listed as 0. You'll change the data classification to the quantile method, which will sort the recorded pressures into even groups that won't be skewed by a few outliers in the dataset. For Counts and Amounts (Color), click Options. Under Classify Data, for Using, choose Quantile.
How accurate were your predictions?. Were there other areas of high or low pressure that you did not predict?. Are there any areas of high or low pressure where you wouldn't expect them to be?. Which areas may soon experience rainfall based on air pressure?. Save the map.Predict rain with temperatureYou've predicted precipitation based on existing rainfall, wind direction, and air pressure. But there are more factors that influence rainfall, including heat and humidity.The amount of water vapor that air can hold depends on its temperature (hotter air holds more).
When air holds the maximum amount of water vapor possible, it becomes saturated. The water vapor begins to condense into tiny water droplets to remove it from the air. This condensation can lead to precipitation.
Saturation can occur when hot air holding a lot of moisture cools suddenly.Your temperature data contains a field called Dew Point Temperature. The dew point temperature is the temperature to which the air would have to cool to become saturated. Thus, the dew point temperature is a measure of how much moisture is in the air. If the dew point temperature is close to the air temperature, the air has a high relative humidity and may soon become saturated. When there is a large difference between the dew point and air temperatures, then the air is dry.To determine where the dew point and air temperature is close, you'll create an Arcade expression that changes the style. You can't create Arcade expressions for sublayers, so you'll first have to add a new version of the original wind speed and direction layer that only shows stations, not buoys.
NOAA METAR current wind speed direction - Stations layer is added to the map. This layer is similar to your NOAA METAR Wind Speed and Direction layer, but it doesn't include sublayers (and only shows stations). Rename the NOAA METAR current wind speed direction - Stations layer to NOAA METAR Dew Point Temperature Difference. Open the Change Style pane for the NOAA METAR Dew Point Temperature Difference layer. For Choose an attribute to show, choose New Expression. A window to create a new Arcade expression appears. The expression you create will be simple.
It will subtract the dew point temperature from the air temperature to find the difference between the two values. Then, it will determine whether that difference is greater or less than 4 degrees Fahrenheit.If the difference is less than 4, saturation—and possibly precipitation—is close. If the difference is greater than 4, saturation is less likely to occur. Next to Custom, click Edit. Change the name of the expression to Dew Point Temperature Difference and click Save.
Under Globals, for Field: Air Temperature, click $feature.TEMP (you may need to scroll down). Note:The number 4 was chosen because it was small, but if the difference is 6 or even 10 degrees, an area might still be at risk of saturation. For an optional challenge, in the Change Style pane, click the Edit Expression button. Change your expression to find values where the difference is less than 6 or 10 degrees and compare your results to when the difference is less than 4 degrees.
What patterns do you find?. What differences exist between the various expressions, and what similarities?. In the Change Style pane, click Done. Turn on the NEXRAD Precipitation layer. Is your dew point temperature difference expression a good predictor of precipitation?.
If you completed the optional challenge (see the note in the previous step), which expression seems to be the best predictor of precipitation?. Based on the dew point temperature difference, which areas aren't currently experiencing precipitation but might soon?It's difficult to compare both the dew point temperature difference and the air temperature at the same time, because the layers mostly overlap. Another way to compare them is to label features. You'll label your dew point temperature difference layer with the air temperature at each point.
In the Contents pane, point to the NOAA METAR Dew Point Temperature Difference layer, click More Options, and choose Create Labels. Does air that is close to being saturated tend to be warmer or colder? Or is there no correlation?. Does precipitation tend to occur more near cold air that is close to saturation or warm air?Examine hurricanesHurricanes are large storms that tend to form over the ocean. They can make landfall and cause property damage and loss of life, and often have strong winds, precipitation, and low pressure. Next, you'll examine your hurricanes layer and see what connection you can find between hurricanes and some of your other data layers.
Based on the Observed Track, what kind of storm did this hurricane begin as? Has it gotten stronger or weaker over time? (In the legend, the symbols are ranked from weakest to strongest, with Hurricane5 being the strongest.). Based on the Forecast Position, what kind of hurricane will this storm become in the next few days? Will it become stronger or weaker than it currently is?Hurricane Alcide started as a relatively weak tropical storm but eventually became a Category 2 Hurricane.
In the next few days, it is forecasted to become weaker, eventually becoming a tropical storm again and not a hurricane. Use the Measure tool to measure how far the hurricane has moved each day in kilometers. Can you see the hurricane in the imagery? How big is it compared to the width of the hurricane track line?.
What is the shape of the cloud cover around the hurricane?. If this hurricane makes landfall, about how much area will it cover?. Return to the Continental United States bookmark. Turn off the Active Hurricanes and GOES Satellite Imagery Transparent layers.Interpolate temperatureSo far, the weather predictions you've made have been about determining what weather will be like in the future.
However, sometimes it's important to make predictions about what the weather is now in areas where no data is available.Your temperature, pressure, and wind speed data all comes from weather stations around the world. But there isn't a weather station covering every location worldwide. How can you know what the weather is in an area with no station? One way is to interpolate a surface.
![Interpolate Points Arcgis Interpolate Points Arcgis](http://fatra.cnr.ncsu.edu/~hmitaso/gmslab/interp/F5f.gif)
Interpolation estimates unknown values across space based on their proximity to known values. Basically, it uses the data you have to make guesses about the data you don't.You'll interpolate the temperature data for a defined geographic area, one with enough data that you can feel confident that your interpolation will be accurate but with enough gaps to make interpolation useful. For this exercise, you'll choose the state of California, in the United States. First, you'll filter your States layer to show only California. The tool runs and the layer is added to the map. Your result will look different than the example image. Is there a relationship between temperature and proximity to the ocean?
If so, what is it?. California has some of the highest and lowest elevations in the United States. Is there a relationship between temperature and elevation? Try using a basemap that shows topographic features such as mountains to help answer the question.
How well does your interpolated surface match the station temperature data?. The Interpolate Points tool uses a statistical interpolation method called. Based on the method's documentation, how confident are you in the accuracy of your interpolated surface?.
For an optional challenge, try creating an interpolated surface for atmospheric pressure or wind speed in California (you can do so by changing the Choose field to interpolate parameter of the Interpolate Points tool). What patterns do you see in the results? How do these patterns differ from the patterns you see in the interpolated temperature surface?. Zoom to the northeastern corner of California. How many weather stations are in this area?. How confident are you in the predicted temperature surface here compared to the San Francisco Bay area? Why?.
Do you think the interpolated surface for this area would be much more accurate if your interpolated surface included data from the weather stations in Oregon and Nevada? Which area or areas in California might have more accurate interpolated surfaces if you included weather stations from neighboring states?. For an optional challenge, try creating an interpolated temperature surface for the states of California, Nevada, Oregon, and Arizona (you can do so by adding expressions to the States layer filter). What differences do you notice?.
Navigate to the North Africa bookmark. If you were to create an interpolated surface of the country of Algeria, how confident would you be in it compared to the interpolated surface you created for California?. Which areas around the world do you think would have the most accurate interpolated surfaces? Which would have the least?. For an optional challenge, try creating an interpolated temperature surface for the country of Algeria. What layer would you need to add to your map to create this interpolated surface? Where can you find this layer?.
Navigate to the California, United States bookmark. Turn on the Counties layer. Counties are subdivisions of states and generally much smaller. About how many weather stations are there per county?. Would you be confident in an interpolated surface that was created for a single county?. Navigate to the Continental United States bookmark. Turn off the States, Counties, and California Temperature Interpolation layers.
Save the map.You've now predicted the weather not only across time, but also space. Throughout this lesson, you've learned how temperature, precipitation, pressure, and wind combine to create the temperature we experience. You answered questions about your data and performed statistical analysis on it to derive new insight.You can apply many of the concepts you learned in this lesson to any weather-related GIS workflow—or even create your own workflow. What questions can you answer using your real-time data that weren't asked in this lesson?You can find more lessons in the.
Summary
Predicts values at new locations based on measurements from a collection of points. The tool uses point data with values at each point as input and makes areas classified by predicted values.
Examples include the following:
![Interpolated Interpolated](/uploads/1/2/6/3/126329206/948769342.jpg)
- An air quality management district has sensors that measure pollution levels. Interpolate Points can be used to predict pollution levels at locations that don't have sensors, such as locations with at-risk populations, for example, schools or hospitals.
- Predict heavy metal concentrations in crops based on samples taken from individual plants.
- Predict soil nutrient levels (nitrogen, phosphorus, potassium, and so on) and other indicators (such as electrical conductivity) in order to study their relationships to crop yield and prescribe precise amount of fertilizer for each location in the field.
- Meteorological applications include the prediction of temperatures, rainfall, and associated variables (such as acid rain).
Usage
-
A point layer is used as the input. The input layer must have a numeric field to serve as the basis of the interpolation. Interpolate Points is designed to work with data that changes slowly and smoothly over the landscape, like temperature and pollution levels. It is not appropriate for data such as population or median income that change very abruptly over short distances.
-
The Interpolate Points tool can be set to optimize speed or accuracy, or a middle ground. The more accurate the predictions, the slower the results take to calculate and vice versa.
-
A layer of standard errors can be created by this tool using the output prediction error option. A 95 percent confidence interval can be calculated for the interpolated layer by taking the interpolation value and adding two standard errors for the upper limit and subtracting two standard errors from the lower limit.
-
This tool uses the Esri Empirical Bayesian Kriging method to perform the interpolation. The parameters that are supplied to this method are controlled by the interpolate option. The parameters are outlined below.
Parameter Speed Default Accuracy Data transformation typeNONENONEEMPIRICALSemivariogram model typePOWERPOWERK_BESSELMaximum number of points in each local model5075200Local model area overlap factor11.53Number of simulated semivariograms30100200Minimum neighbors81015Maximum neighbors81015
Syntax
Parameter | Explanation | Data Type |
The point features that will be interpolated to a continuous surface layer.
|
Feature Set | |
outputName
|
The name of the output layer to create on your portal.
|
String |
(Optional)
|
The numeric field containing the values you want to interpolate.
|
Field |
(Optional)
|
Controls your preference for speed versus accuracy, from fastest to most accurate. More accurate predictions take longer to calculate.
|
String |
(Optional)
|
If checked, a polygon layer of standard errors for the interpolation predictions will be output.
Standard errors are useful because they provide information about the reliability of the predicted values. A simple rule of thumb is that the true value will fall within two standard errors of the predicted value 95 percent of the time. For example, suppose a new location gets a predicted value of 50 with a standard error of 5. This means that this task's best guess is that the true value at that location is 50, but it reasonably could be as low as 40 or as high as 60. To calculate this range of reasonable values, multiply the standard error by 2, add this value to the predicted value to get the upper end of the range, and subtract it from the predicted value to get the lower end of the range.
|
Boolean |
(Optional)
|
Determines how predicted values will be classified into polygons.
|
String |
(Optional)
|
This value is used to divide the range of predicted values into distinct classes. The range of values in each class is determined by the classification type. Each class defines the boundaries of the result polygons.
The default is 10 and the maximum is 32.
|
Long |
[classBreaks,...]
|
To do a manual classification, supply the desired class break values. These values define the upper limit of each class, so the number of classes will equal the number of entered values. Areas will not be created for any locations with predicted values above the largest entered break value. You must enter at least 2 values and no more than 32.
|
Double |
boundingPolygonLayer
|
A layer specifying the polygons where you want values to be interpolated. For example, if you are interpolating densities of fish within a lake, you can use the boundary of the lake in this parameter and the output will only contain polygons within the boundary of the lake.
|
Feature Set |
predictAtPointLayer
|
An optional layer specifying point locations to calculate prediction values. This allows you to make predictions at specific locations of interest. For example, if the input layer represents measurements of pollution levels, you can use this parameter to predict the pollution levels of locations with large at-risk populations, such as schools or hospitals. You can then use this information to give recommendations to health officials in those locations.
|
Feature Set |
Derived Output
Name | Explanation | Data Type |
outputLayer |
The output polygon features, where each polygon surrounds interpolated values based on the classification type and number of classes.
|
Feature Set |
outputPredictionErrorLayer |
Contains the predicted error for each point in the input layer.
|
Feature Set |
outputPredictedPointsLayer |
The point layer containing points from the predicted point layer with their predicted values.
|
Feature Set |
Environments
Licensing information
- Basic: Requires your account in ArcGIS Enterprise to have the Perform Analysis privilege
- Standard: Requires your account in ArcGIS Enterprise to have the Perform Analysis privilege
- Advanced: Requires your account in ArcGIS Enterprise to have the Perform Analysis privilege
Related topics
![](/uploads/1/2/6/3/126329206/193113772.jpg)