Check the behavior of the product pools

[luyssaert]

Following LCC the product pools are empty. This could be do to the fact that LCC now takes place the first day of the year and that the product pools are calculated the last day of the year. In between the inout of product use are set to zero. Along the same line there seems to be a problem with the spatial distribution of burried_c

[luyssaert]

• LCC is now calculated the first day of the year, wood products were still calculated the last day of the year. This inconsistency was removed in r6329.

[luyssaert]

The call sequence of product_init was changed and the flag do_now_product use was added. This flag is always true the first day of the year. This flag cannot be set by the user. Note that product use is calculated for both wood harvest from forest management and wood harvest from land cover change (LCC). This is one of the reasons why it should be calculated alos in the absence of LCC.

r6358 was run for 5 years with LCC and product pools started to develop in y1-y5. When the model was run without LCC no product pools developed (y1-y5). Subsequently the run without LCC was restarted with LCC and product pools started to develop (y1-y5 but y6-y10 remained empty). Finally, the LCC run was restarted without LCC. Now the product pools got older but no new wood from LCC was added (y1-y5 remained empty but products of age 6 to 10 remained). The pools now behave as expected.