Unité de recherche en sol et systèmes d’informations spatiales, RUSSIS, Faculté de Gestion des Ressources Naturelles Renouvelables, Université de Kisangani, BP 2012, Kisangani, RD Congo
The different forest ecosystems of the RDC have been affected by the dynamics of the soil occupation during these last decades, characterized by a clean reduction of the natural formations to the profit of the formations anthropiques. The studies achieved on this day limited themselves to an analysis diachronique by the integrated approaches implying the techniques of spatial analysis combined to the geographical information system (SIG).
Setting to part the analysis diachronique of the imagery LANDSAT TM, ETM+ and OLI (of parth and Raw 177-59, 1766-59 and 176-60) spreading of years 1986 to 2015 and supported of direct observations, the present survey sketches a prospective analysis of this dynamics while combining the applicable variables of the changes and the deterioration of the forests to a test of modelling to the deadline 2065.
The variables have been regrouped in 7 subgroups: agriculture, lumbering, infrastructures, the demographic factors, the socio-politic factors, the economic factors and the biophysical factors. Nevertheless, only the variables spatially explicit have been quantified and steady to the models. For the rest, two tools of modelling (CA_Markov and LCM) have been used and the best allowed of cartographier the risks of deforestation the deadline 2065; the modelling has been made to the scale of the reserve of biosphere of Yangambi. For the two models a based objective and quantitative approach on the budgeting of the mistakes and the correct predictions has been adopted finally.
The results show a meaningful regression of the forest formations to the profit of the classes anthropisées with rates of deforestation of 1986-2002 and 2002-2015 of 1,1% and 4,31% and respectively. These reports come proved the necessity of the setting in application of a planning plan in view of the lasting management.
The visual and statistical comparison of the affected and real cards of 2015 allowed to use the tool THAT Markov for the simulation of the variation in surface of the classes of soil occupations the deadline 2065. However, the omen of an alarming and continual regression if some consequent measures are not set in motion.