Adapted methods of spatial information analysis developed and implemented in the framework of IACS-LPIS setup in Romania.

Docan, Daniela Cristiana ; Dana, Iulia ; Badea, Alexandru 等

Abstract: Romanian Paying and Intervention Agency for Agriculture has financed a complex project in order to create and maintain the IACS-LPIS (Integrated Administration and Control System--Land Parcel Identification System). The orthophotos needed for the creation of LPIS were provided by National Agency for Cadastre and Land Registration (ANCPI). The multidisciplinary quality control team, in charge with the realization of the project has developed innovative procedures based upon ESRI technologies (VISAROM, ArcGIS and GIS DataReViewer). These procedures were customized and tested in order to fulfill the specific requirements of Romania. Spatial data quality control (QC) tasks include a series of automated and visual checks of the orthophotos, database, topology and accuracy of the physical blocks digitization.

Keywords:orthophotos,GIS,database,topology,quality control.

1. INTRODUCTION

The application presented in this paper was realized in the framework of LPIS setup in Romania. LPIS has the goal to implement the agricultural subsidies, according to the policy and procedures of the European Union (EU). The project was financed by Romanian Paying and Intervention Agency for Agriculture (APIA). The abbreviation LPIS is used by European Union in order to define the Land Parcel Information System as a key component of the Integrated Administration and Control System (IACS) for area-based subsidies, developed by all the European countries (member or on road to become members) according to the fundamental juridical regulations of UE. IACS is mainly used for the management and the control of agricultural subsidies by all EU national administrations. LPIS holds information about land parcels which are used as a reference for declaration by farmers, for administrative and cross checks (100% of all declarations) and finally for on the spot inspections (minimum 5% of all declarations) (Badea et al.,2004).

2. SPATIAL DATA ORGANIZATION

According to the requests and recommendations of UE, the creation of LIPS has to take into account the following aspects:

-utilization of orthophoto

-creation of the database in a GIS environment

-homogeneity of the database

The established GIS is based on color digital orthophotos (pixel size 0.5m, accuracy [+ or -]1.5m) created from aerial photographs acquired thought 7 different aerophotogrammetric projects between 2002-2006, delivered by the National Agency for Cadastre and Land Registration. The orthophoto database consists of more than 39000 tiles having a size of 2.5kmx2.5km. In order to facilitate the access to image information, it was decided (together with the beneficiary of the project) to organize these tiles in a configuration of 10kmx10km. This process involved the mosaicking of the approximately 2.5kmx2.5km tiles resulting 2537 tiles with the dimension of 10kmx10km, in a geotiff format.

[FIGURE 1 OMITTED]

These 10kmx10km tiles were compressed in MrSID format (second version) in order to be compatible with the IT system based on LaFIS--software designed by GAF for administration purposes granting completely the current EC guidelines.

In order to receive the agricultural subsidies, the farmers have to declare their intention to be beneficiary and must identify the land parcel used for agricultural purposes, in the area of a "physical block", also known as the "reference parcel" (APIA,2006).

The digitization of physical blocks based on orthophoto maps (supplied by ANCPI and corresponding to the Romanian administrative units) constituted the first step of the project. The next step consisted in the elaboration of the spatial database necessary for an easy identification of the parcels by the farmers

In order to have a unique and continuous cartographic representation for the entire Romanian territory, the national projection Stereographic '70 was selected. For filling in the GIS database concerning the limits of the administrative units, a *.shp file, provided by ANCPI, was used. The minimum accuracy of the database has to be in accordance with the cartographic standards for a scale of 1:10.000. The use of the 0.5m resolution orthophoto maps and the process of digitization at a higher scale (1:1000-1:2000) has automatically guarantied this condition.

The QC team agreed with the beneficiary (APIA) to operate specific recommendations and to propose solutions for each of the named activities in order to homogenize and to make compatible the activity of contractors.

[FIGURE 2 OMITTED]

3. QUALITY CONTROL OF DATA

The external quality control was based on two complementary actions which determine if the work of contractors is according with the technical specifications concerning the:

-respect of data format

-quality of photo interpretation--the unit for quality control was the administrative unit (UAT)

3.1. Phase I

Visa-Rom application developed by Geosys France following the specifications elaborated by CRUTA and ROSA was used for the quality control of data delivered by the contractors in charge with the digitization of the physical blocks. The functions implemented in this application allow an accurate analysis of the data by identifying the possible topological errors (overlaps, multipart polygons, and gaps), the accuracy of the database fields, according to the European standards.

The first control is realized (topology and structure of data) using Visa-Rom (Map Objects based software application) for 100% of UAT's surface. The check of photo interpretation is realized for a 5% sample of physical blocks (sample generated by Visa-Rom).

The administrative unit is considered REJECTED if more than 20% of the 5% physical block sample is presenting errors. The commune rejected is retransmitted to the contractor to correct the digitization. If the number of rejected blocks is less or equal with 20% of the 5% sample, the administrative unit is considered ADMITTED.

3.2. Phase II

A sample of 30 communes has been chosen in order to test, set up the final checks (quality control), and identify/correct the database errors. The goal was to identify geometrical and attribute errors and to correct the attribute errors. The control was realized for 100% of the physical blocks in an administrative unit.

The existing data for each commune were:

* the limits of the commune in shape file format;

* the physical blocks digitization by contractors;

* the orthophotos at 0.5m resolution in MrSID format.

In order to perform the QC it is necessary to be loaded in an intermediated workspace (a Personal GeoDatabase which allows to perform a series of compulsory checks). Practically, each commune was loaded in a distinct Geodatabase.

Geodatabase topology rules are used as initial quality control checks for spatial relationships such as adjacency, coincidence and connectivity (Zeiler&Environmental Systems Research Institute,Inc.,1999).

3.3.Perform automated QC checks using GIS Data ReViewer

GIS Data ReViewer provides tools for selecting and editing, performing spatial analysis, finding overlapping features, creating error tables, capturing and digitizing missing features, and making topology checks. Errors are logged into a table that stores all the necessary information for historical tracking (Environmental Systems Research Institute, Inc. 2006).

3.4.Perform visual QC checks using GIS Data ReViewer

This step assumes visual checks of each physical block in order to identify photointerpretation and attribute errors and to correct only the attribute errors.

[FIGURE 3 OMITTED]

[FIGURE 4 OMITTED]

4. CONCLUSION

The intelligible and detailed technical approach leads to a homogeneous and conform evaluation of the digitization/photointerpretation activity.

The training of the operators in charge with the digitizing process by photo interpreting is an important element for a final quality product. In forest area, due to excessive digitization of the boundaries, was generated a high number of vertices at distances lower than 1 meter. That is affecting the size of the database. It is necessary to underline the fact that the final QC process, didn't allow the modification of the number of existing polygons (physical blocks) because the parallel process of farmers declarations is referring to the initial form of the database.

In fact, the adopted scenario of photo interpretation and the correct definition of the database structure have a decisive impact on the quality of the operational database. Finally, this database has to support the long distance networking access for more than 2000 concurrent users allowing them to perform the update of the content accordingly the IACS/LPIS requirements.

5. REFERENCES

APIA(2006).The APIA guide to IACS-LPIS create in Romania

Badea, A & Killmayer, A(2002). First Pilot Operation for the Constitution of a Land Parcel Information System in Romania, 8th Annual CwRS Conference, Varese, Italy, Available from:http://mars.jrc.it/documents/control/ meetings/varese2002/proceedings/session3/S3_Badea.JPG

Badea, A; Killmayer, A & Nedelcu, I. (2004). Pilot project and Strategy for the implementation of a Land Parcel Identification System in Romania, 10th Annual Conference on Control with Remote Sensing of Areabased Subsidies,25-27th of November, Budapest, Hungary

Environmental Systems Research Institute, Inc. (2006). QA/QC for GIS Data

Zeiler,M&Environmental Systems Research Institute,Inc. (1999). Modeling our world-The ESRI Guide to geodatabase Design