Application of sewage sludge compost (SSC) as a fertilizer on landscaping provides a potential way for the effective disposal of sludge. However, the response of landscape trees to SSC application and the impacts of heavy metals from SSC on soil are poorly understood. We conducted a pot experiment to investigate the effects of SSC addition on Mangifera persiciforma growth and quantified its uptake of heavy metals from SSC by setting five treatments with mass ratios of SSC to lateritic soil as 0%:100% (CK), 15%:85% (S15), 30%:70% (S30), 60%:40% (S60), and 100%:0% (S100). As expected, the fertility and heavy metal concentrations (Cu, Zn, Pb and Cd) in substrate significantly increased with SSC addition. The best performance in terms of plant height, ground diameter, biomass and N, P, K uptake were found in S30, implying a reasonable amount of SSC could benefit the growth of M. persiciforma. The concentrations of Cu, Pb and Cd in S30 were insignificantly different from CK after harvest, indicating that M. persiciforma reduced the risk of heavy metal contamination of soil arising from SSC application. This study suggests that a reasonable rate of SSC addition can enhance M. persiciforma growth without causing the contamination of landscaping soil by heavy metals.

Forests suffer from heavy losses due to the occurrence of fires. A prediction model based on environmental condition, such as meteorological and vegetation indexes, is considered a promising tool to control forest fires. The construction of prediction models can be challenging due to (i) the requirement of selection of features most relevant to the prediction task, and (ii) heavily imbalanced data distribution where the number of large-scale forest fires is much less than that of small-scale ones. In this paper, we propose a forest fire prediction method that employs a sparse autoencoder-based deep neural network and a novel data balancing procedure. The method was tested on a forest fire dataset collected from the Montesinho Natural Park of Portugal. Compared to the best prediction results of other state-of-the-art methods, the proposed method could predict large-scale forest fires more accurately, and reduces the mean absolute error by 3–19.3 and root mean squared error by 0.95–19.3. The proposed method can better benefit the management of wildland fires in advance and the prevention of serious fire accidents. It is expected that the prediction performance could be further improved if additional information and more data are available.