Difference between revisions of "Productivity Maps"

From Wikiagro.com
Jump to: navigation, search
Line 5: Line 5:
 
This is a procedure relatively simple, automated and objective, that allows to compare the field production and variability in different times, in a single map. Remote sensing imagery is selected in key dates, when crops are vigorous, and Vegetation Index (NDVI) maps are generated on those dates. These layers are then normalized, and integrated in the management zone map.  
 
This is a procedure relatively simple, automated and objective, that allows to compare the field production and variability in different times, in a single map. Remote sensing imagery is selected in key dates, when crops are vigorous, and Vegetation Index (NDVI) maps are generated on those dates. These layers are then normalized, and integrated in the management zone map.  
  
 
+
<br>
  
 
=== Methodology - Cluster analysis  ===
 
=== Methodology - Cluster analysis  ===
Line 11: Line 11:
 
A cluster is a predefined area, for statistical analysis purposes. For instance a cluster could be the pixel of a satellite image, such as 30x30 mts for Landsat images, 15 x 15 mts for Aster images, or 10 x 10 mts for Spot images. Once the Cluster area is defined, a normalized NDVI map is generated, with a grid with the cluster size.  
 
A cluster is a predefined area, for statistical analysis purposes. For instance a cluster could be the pixel of a satellite image, such as 30x30 mts for Landsat images, 15 x 15 mts for Aster images, or 10 x 10 mts for Spot images. Once the Cluster area is defined, a normalized NDVI map is generated, with a grid with the cluster size.  
  
the following sample images and corresponding NDVI illustrate the process:
+
the following sample images and corresponding NDVI illustrate the process:  
  
'''//insert images in the table'''
+
'''//insert images in the table'''  
  
 
{| cellspacing="1" cellpadding="1" border="1" width="200"
 
{| cellspacing="1" cellpadding="1" border="1" width="200"
 
|-
 
|-
|  
+
| <br>
|  
+
| <br>
 
|-
 
|-
|  
+
| <br>
|  
+
| <br>
 
|-
 
|-
|  
+
| <br>
|  
+
| <br>
 
|}
 
|}
 +
 
[[Image:IMAGEN 1.jpg|Image:IMAGEN_1.jpg]] Imagen Landsat 25-01-2006  
 
[[Image:IMAGEN 1.jpg|Image:IMAGEN_1.jpg]] Imagen Landsat 25-01-2006  
  
Line 32: Line 33:
 
[[Image:IMAGEN 3.jpg|300x200px]] Imagen Landsat 08-03-2004  
 
[[Image:IMAGEN 3.jpg|300x200px]] Imagen Landsat 08-03-2004  
  
 +
<br>
  
 
+
These images are integrated by analyzing each cluster, defining homogeneous production areas, and obtaining the management zones map.  
These images are integrated by analyzing each cluster, defining homogeneous production areas, and obtaining the management zones map.
+
  
 
'''//insert Image:'''Mapa_productividad3.jpg  
 
'''//insert Image:'''Mapa_productividad3.jpg  
 +
 +
=== <br>Limitations ===
 +
 +
Processing is carried out at the field level. The main reason is that in a given farm, fields may have at the same time different types of crops or uses. Thus, NDVIs from different fields cannot be compared. For this reason different images are selected for different fields.
 +
 +
As a conclusion, management zones maps should only be used to compare variability within fields, but not as a means of comparison between different fields.
 +
 +
=== Applications and Benefits ===
 +
 +
Management zones maps provide guidance to know the variability within a given field. Based on this knowledge, field visits can be carried out, making decisions regarding field management, management zone definition, selecting areas for tests, etc.
 +
 +
Using Management zones to define sampling points makes possible to review on site the different zones and the reasons behind the possible limitations.
 +
 +
This can be a starting point to define inputs and recommendations, selecting the type and quantity of inputs according to the management zones.<br>
  
 
<br>
 
<br>

Revision as of 22:02, 24 August 2009

Overview

Management Zones maps are generated based on the combination of Vegetation Index maps, through a process called 'cluster analysis'. It's also possible to include information from previous yield monitor maps if available. 

This is a procedure relatively simple, automated and objective, that allows to compare the field production and variability in different times, in a single map. Remote sensing imagery is selected in key dates, when crops are vigorous, and Vegetation Index (NDVI) maps are generated on those dates. These layers are then normalized, and integrated in the management zone map.


Methodology - Cluster analysis

A cluster is a predefined area, for statistical analysis purposes. For instance a cluster could be the pixel of a satellite image, such as 30x30 mts for Landsat images, 15 x 15 mts for Aster images, or 10 x 10 mts for Spot images. Once the Cluster area is defined, a normalized NDVI map is generated, with a grid with the cluster size.

the following sample images and corresponding NDVI illustrate the process:

//insert images in the table







Image:IMAGEN_1.jpg Imagen Landsat 25-01-2006

279x200px Imagen Landsat 06-01-2005

300x200px Imagen Landsat 08-03-2004


These images are integrated by analyzing each cluster, defining homogeneous production areas, and obtaining the management zones map.

//insert Image:Mapa_productividad3.jpg


Limitations

Processing is carried out at the field level. The main reason is that in a given farm, fields may have at the same time different types of crops or uses. Thus, NDVIs from different fields cannot be compared. For this reason different images are selected for different fields.

As a conclusion, management zones maps should only be used to compare variability within fields, but not as a means of comparison between different fields.

Applications and Benefits

Management zones maps provide guidance to know the variability within a given field. Based on this knowledge, field visits can be carried out, making decisions regarding field management, management zone definition, selecting areas for tests, etc.

Using Management zones to define sampling points makes possible to review on site the different zones and the reasons behind the possible limitations.

This can be a starting point to define inputs and recommendations, selecting the type and quantity of inputs according to the management zones.