Forest carbon dynamics associated with growth and disturbances in Oklahoma and Texas, 1992-2006

Quantifying forest carbon changes associated with growth and major disturbances is important for management of greenhouse gas emissions related to forests. Regional-level approaches with improved local growth data may refine estimates obtained using coarser resolution information. This study integrates remote-sensing-derived land cover change products, harvest data, forest fire data, and local forest growth estimates at the county level to identify forest ecosystem carbon change for the states of Oklahoma and Texas (1992-2006). Whereas Oklahoma was a carbon sink of 0.5 Tg C yr-1, Texas was estimated to be a carbon source of -1.8 Tg C yr-1 for the period. The two states together functioned as a carbon source of -1.3 Tg C yr-1 for the entire period, although it was a small sink of 0.1 Tg C yr-1 in the recent period of 2001-2006 due to reduced annual rates of net forest-to-nonforest conversion and harvesting, compared to those in the early period of 1992-2001. Most counties located in the western portions of both states were small sinks of carbon during the period. Even though their growth rates are greater, many counties in the eastern portions of both states were carbon sources due to a higher intensity of forest-related disturbances. A sensitivity analysis was conducted to investigate possible double-counting of harvest and cover change by assuming half of the sequestration and emissions from land cover changes were already counted as harvest. Results indicated Oklahoma would be a sink of 1.0 Tg C yr-1, and Texas would be a small carbon source of -0.1 Tg C yr-1. Uncertainty in forest area for the western portions of these states remains an important source of potential error.