Pasture area comparison

Progression of pasture area through time.

Pasture area for Lapig and MapBiomas in 2017 using the same breaks

The differences between the two sources of information is very low, except for Corumba in Mato Grosso do Sul. In this case MapBiomas is considering the biggest part of the area as grassland instead of pasture.

The total pasture area is not enough to explain the difference in cattle deforestation observed in trase-earth, this is due to the method used in cattle deforestation (at the present moment), that is considering the increment of the area instead of the total area or the method considering the 5 years allocation period. In the figure below the increment of the area in 2017 in relation to 2016. This is also not exactly what we are using in trase.earth (increment from the mask), but more closest. Values lower than 0, which means a reduction of pasture between 2016 and 2017 were removed from the palette once the reduction of pasture is not possible in our method.

The major part of the “increment” of pasture in Lapig data sizes from 0 to 5000 ha or up to 50km2. Municipalities without strong agricultural activities, consequently less susceptible to lose their pasture areas in the 5-year windows to another commodity, show much lower deforestation in Lapig than in MapBiomas. This is the opposite than expected by other sources like Prodes, where this kind of municipalities exhibit deforestation around 4x higher than presented by Lapig
Prodes: Apui-AM (169km2), Sao Felix do Xingu-PA (237km2), Altamira=PA (223km2), Colniza-MT(174km2)