# Convert a lat, long, value table into a raster

I would like to to convert a table with latitude, long, and a value into a raster. My file looks like this:

X   Y   Z
32.85138942 -6.643054718    0
32.85138942 -7.059721474    1004
32.85138942 -7.143054826    1014
32.85138942 -7.226388177    1080
32.85138942 -7.309721529    1040
32.85138942 -7.39305488     0
32.85138942 -7.476388231    971
32.85138942 -7.559721583    0
32.85138942 -7.643054934    3092

I tried the following:

1) Using the Point to Raster tool

a. importing the CSV file into ArcGIS
b. displaying the XY data
c. saving it as a shapefile
d. using the “Point to Raster tool” to turn the shapefile into a raster

However, this only give me a series of squares that are not adjacent together, and not a continuous surface.

2) Using the XYZ2GRID2.1-Batch program; but I can't get it to work.

Is there an other way to convert this type of lat, long, values into a raster?

• When you used the 'Point to Raster' tool, what did you specify as the 'Cellsize'? You appear to have a gridded data set with a cellsize of 0.08333 degree. If assigning 0.083333 as the cellsize doesn't help, you could also try to use a interpolation method. – Mikkel Lydholm Rasmussen Mar 18 '15 at 15:19
• When opened in notepad++ are the records like: 32.85138942 -6.643054718 0 each on their own line? – HeikkiVesanto Mar 18 '15 at 15:41
• This kind of "conversion" is called interpolation because it has to create data where none yet exist. Please research the interpolation related threads on this site to learn more. How you interpolate depends on what these data mean. If they are point observations of a true underlying continuous surface, then the interpolation must predict the unobserved values. If instead they are observations of a population of values (such as numbers of people, houses, trees, etc), then you probably want a density estimator. – whuber Mar 18 '15 at 20:23
• Would your ideal end product look like a raster of the point data extent with individual pixels either 0 (background) or the value in the 3rd column of your csv? – Aaron Mar 18 '15 at 20:24