Interpolating Rainfall from Point Data to Raster Layers using Inverse Distance Weighting (IDW)

This blog details how to convert precipitation (rainfall) data obtained from weather stations into shapefiles (point data) before interpolating the shapefiles into raster layers on ArcMap using the IDW tool.

Here we use the total annual precipitation over 8 years from three weather stations to interpolate over the area between the points, producing a mean total annual precipitation raster layer.

Converting Raw Data To Shapefiles

1. Organise the raw data on Excel, providing spatial information (coordinates of the weather stations) and save it as a .csv file (so it can be read on ArcMap). Key columns include station name, X and Y coordinates of each station, year of data collection and precipitation (mm) [Table 1].

2. Open the .csv file on ArcMap and display the XY coordinates of the data (Right Click on the layer > Display XY Data > Identify columns that contain XY data and set the coordinate system). This should display the geographic coordinates of the three weather stations as three individual points for the years 2010 to 2018.

3. Export these points to a new shapefile layer (Right Click on the layer > Data > Export Data) and save it in a designated workspace. You have now converted your raw precipitation data into a shapefile layer [Figure 1].

Table 1: Key columns when organising the raw precipitation data.

Note: Although there are 8 different values of total annual precipitation in each station (as the data spans over 8 years), the coordinates for each year is identical in each station. Hence, what is 8 individual point values for each station is displayed as one point per station due to points overlapping each other.

Figure 1: Displayed XY coordinates of the raw data.

Interpolating Shapefiles To Raster Using IDW

1. Import the shapefile onto ArcMap.

2. Using the IDW tool (ArcToolbox > Spatial Analyst Tools > Interpolation > IDW)

3. Under ‘input point features’, select the shapefile and select ‘Precipitation’ under the ‘Z value field’. Name the output raster appropriately and select a workspace to save it to. To use the optional fields, click on ‘Show Help’.

4. In the ‘Environment Setting’, set the desired coordinate system under ‘Output Coordinate System’. Click ‘OK’ when done.

5. Your shapefile precipitation data is now converted into a raster layer [Figure 2].

Tip: If you would like to extrapolate the raster to cover areas outside the perimeter of your points, specify the extent to which precipitation should be extrapolated under ‘Processing Extent’ in Step 4 [Figure 3].

Figure 2: Interpolated raster image of the precipitation shapefile.

Figure 3: Extrapolated raster image of the precipitation shapefiles.

Limitations: The data set used in this guide is rather small for accurate interpolation over the study area. Statistical significance generally requires a minimum of 30 points per variable. However, the more points you have, the more accurate your IDW raster is likely to be.

There are other methods of interpolating point data into raster and it is important to choose a method based on the type of data you are working with and its purpose.

The link below leads to a paper on three different methods of interpolating precipitation data from rainfall stations:

Spatial Interpolation of Annual Precipitation in Annaba-Algeria - Comparison and Evaluation of Methods. Available at: https://www.researchgate.net/publication/257711970_Spatial_Interpolation_of_Annual_Precipitation_in_Annaba-Algeria_-_Comparison_and_Evaluation_of_Methods [Accessed May 14 2019].

---

Cover image from https://azclimate.asu.edu/weather/state-tempprecip/