Tree belts (TBs) as one of the agroforestry types, provide many ecosystem services (ESs), that reduce the negative environmental impact of agriculture and increase agricultural productivity. Unfortunately, the lack of quantitative and spatial information regarding TB properties, such as the porosity expressed by gap fraction, often leads to insufficient information on the availability of ESs provided by TBs, such as windbreak efficiency, barrier effectiveness for soil erosion and ability to redistribute snow cover. Moreover, image data, which are usually used for the assessment of TB porosity, only provide 2D information. Therefore, we propose the first comprehensive method of TB porosity indicator estimation conducted by gap mapping using 3D LiDAR (Light Detection and Ranging) data. The method allows to diagnose the spatial structure of TBs necessary for reliable evaluation of ES availability. Detailed cross-section diagnostic graphs and TB silhouette maps showed the intra-TB variation of porosity and thus may support the decision making on the crop types and may lead to better predictions of agricultural production. Also, on a testing study area, we revealed that the proposed method is suitable to significantly improve the accuracy of vegetation volume estimation in TB. Such a detailed analysis of the inner TB structure may be used to maximize the positive effect of tree belts on agricultural production and therefore is a way towards the sustainable agriculture as well as better assessment of ES provided by tree belts.