Crop Production Model

[dave]

Hi @anjaney.singh , I’ll try to address each question,

@anjaney.singh

• Does the LULC map for the inputs need to contain the crop detail or other features, such as forestry, barren land, etc?

From the User’s Guide: " Land-Use/Land-Cover Map (required). Raster of land use/land cover (LULC) for each pixel, where each unique integer represents a different land use/land cover class. These integers are used in the Landcover to Crop Table to map landcover classes to specific crops."

So the LULC values should represent crops, and the “landcover to crop table” defines that mapping.

@anjaney.singh

• With regard to the percentile and regression models, what is the output and what does it tell us?

This User’s Guide section describes the results. http://releases.naturalcapitalproject.org/invest-userguide/latest/crop_production.html#interpreting-results
And this thread discusses interpreting results also. Primary output of the model Crop Production (result_table_[Suffix].csv)

Please ask a more specific question if you have any.

@anjaney.singh

• The percentile and regression yields also give results for crops not grown in the area of interest. Is the yield available for all crops in the dataset and what is its relation to the LULC input?

Could you describe or share the outputs you are looking at? I believe the results for each crop should be masked to the LULC pixels representing that crop. And the yield elsewhere will be 0 or nodata. Results should be given for only the crops listed in the “landcover to crop” input table.