The Evaluation Of The Value Of Forest Carbon Fixation And Oxygen Release And Water-conservation Based On Cloud Computing

Cloud computing is the main direction of current computer research.It integrates advanced technologies such as grid computing,parallel computing,and distributed computing,and it is widely used in many industries.Cloud computing forms a resource pool by integrating resources and provides services to users with high availability and scalability.Hadoop is a well-developed cloud computing platform and a mature open source software framework that processes data in parallel has been favored by various industries due to its high reliability and scalability.Advanced satellites and sensors are continuously put into use,making the spatial data scale exponentially growing,and the types of remote sensing data are also constantly enriched.In addition,increasingly complex algorithms and models are gradually being applied to remote sensing data processing,so remote sensing data processing presents data-intensive and computationally intensive features.At present,most remote-sensing processing software is based on a single-machine system and is subject to great restrictions in terms of calculation speed and efficiency,and cannot meet the dynamic monitoring and evaluation of the ecological environment.The application of cloud computing technology to ecological assessment can increase the processing speed of remote sensing images and achieve dynamic disaster prediction and ecological environment evaluation.This paper applies cloud computing technology to forest ecosystem function assessment and remote sensing image processing,and uses the core components of the open source cloud computing platform Hadoop to parallelize the ecological factors of the Maoershan research area.The evaluation of forest ecosystem service functions proves the effectiveness of applying cloud computing to the evaluation of forest ecosystem service functions.The main research contents of this article are as follows:(1)Data storage based on HDFS file system and HBase.In this paper,remote sensing image data is stored in HDFS,and the storage strategy of the image in HDFS is specified;the file storage structure of HDFS is designed,and the data includes original data,preprocessed data,and remote sensing inversion result data.HBase is a column-oriented,unstructured database.In the paper,land cover type data,soil data,and weather data are regarded as text data,and all text data are stored in HBase.HDFS does not directly support the processing of remote sensing images,but provides an extensible API.This paper focuses on the storage and processing of remote sensing images and extends the related functional modules on the basis of Hadoop.After the MapReduce program is executed,data is saved back to HDFS.(2)Remote sensing parameter inversion method based on MapReduce parallel framework.Most remote-sensing processing software is based on a single-machine system,and it is greatly constrained in calculation speed and efficiency.This paper builds a Hadoop-based cluster environment and integrates remote sensing parameters inversion model algorithms into a parallel framework based on MapReduce.A remote sensing inversion algorithm for ecological factors based on the MapReduce parallel framework was constructed to improve image processing efficiency.(3)Remote sensing estimation model(Cloud-ICASA)for Vegetation Net Primary Productivity.In the assessment of the service function of carbon fixation and oxygen release,the parameter(the net forest stand productivity(NPP))is difficult to obtain directly,which is the difference between the carbon absorbed by plants and carbon released by autotrophic respiration during photosynthesis.NPP directly reflects vegetation Production capacity under natural environmental conditions plays an important role in global change and carbon balance.This paper corrects and optimizes the key parameters of the CASA model,and proposes an improved Cloud-ICASA model for Maoershan.(4)Estimation of stand evapotranspiration by Cloud-ISEBAL.In the evaluation of conservation water service functions,the main parameter is the evapotranspiration of the forest.Evapotranspiration is the intermediate link between groundwater and surface water,which affects the water cycle and carbon cycle of the forest ecosystem.With the development of remote sensing technology,the use of remote sensing methods to perform inversion of surface energy becomes a reality.Based on SEBAL,this paper optimizes its key parameters and proposes an improved Cloud-ISEBAL model to achieve remote sensing inversion of forest evapotranspiration.