**3.2. Instruments**

A monitoring station was installed in the north of the Calakmul Biosphere Reserve in the town Modesto Angel (MA) and at the South close to Ley de Fomento town was an Automatic Weather Station [36]. Soil moisture was measured using the direct gravimetric method and also continuously using indirect methods based on reflectometry: time domain (TDR) and frequency domain (FDR). Additionally, other measured variables at this station were physical characteristics of the soil, rainfall, air temperature, and relative humidity.

collect representative data of each of them and in this way observe the distribution of soil moisture in different climatic regimes. The rainy season occurs between the months of June and July, until October and the dry season between December and April or May. During the visits to the study area, the physical condition of the equipment and the environment was recorded. Fieldwork consisted of the installation of equipment, acquisition of soil moisture, vegetation, and meteorological data, which was done every fortnight, period corresponding to the data collection with the diviner sensor. Soil samples were taken for the measurement of gravimetric humidity. Soil samples were of approximately 300 g and obtained in each of the eight sites, and the following data were obtained: soil moisture and physical properties (textural fraction, bulk density, permanent wilting point (PMP), field capacity (CC), electrical conductivity and pH). Subsequently, samples of 100 g were taken every 10 cm in the soil profile to perform the gravimetric procedure and define the amount of gravimetric water content of each of the study sites. The physical characteristics of the soil samples such as bulk density, PWP and CC, electrical conductivity and pH were carried out in the National Forestry, Agriculture and Livestock Research Institute (INIFAP) and in the Soils and Plants Laboratory of the Academic Division of Agricultural Sciences of the Autonomous Juarez University of Tabasco. Once the humidity values of the indirect measurements in situ have been validated, the analysis of their spatial and temporal distribution is carried out making use of geographic information systems and other computer programs for the graphic modelling of the data. The vertical analysis allows the visualisation of the fluctuation of soil moisture for each site, taking into account the relationship with the textural fraction of the soil. The results of this analysis permit the understanding of the mechanism of infiltration, drainage and saturation in the first meters of the soil layer. The temporal resolution to obtain one measurement varies for each technique. The highest temporal resolution can be provided by the TDR and FDR-Decagon (FDR\_Dec) with one observation for every 20 min, the FDR-Diviner 2000 (FDR-Div) can record one measurement for every week, and the gravimetric method can be used for every 4 months. This indicates that one can have more frequent TDR and FDR-Dec observations than the other FDR techniques.

Correlation between TDR and FDR Soil Moisture Measurements at Different Scales to Establish…

http://dx.doi.org/10.5772/intechopen.81477

63

Once measurements with sufficient support at the local scale are obtained and the spatial and temporal stability are established, they can be scaled. Scaling up soil moisture is divided into

meteorological and climatological effects such as precipitation or solar radiation [3]. In this

In order to estimate the accuracy between the three soil moisture methods, a comparison analysis was performed. Statistical indicators such as the coefficient of determination (R<sup>2</sup>

root mean square error (RMSE), relative error, mean bias error (MBE) and normalised root

*n* ∑ *t*=1 *N*

\_\_\_\_\_\_\_\_\_\_\_\_\_\_ \_\_1

(*Di* − *Obsi*

heterogeneity and changes in soil cover; and regional scale, from 50 to 400 km<sup>2</sup>

paper, a small scale is presented since the radium of influence is less than 50 km.

, affected by variations in soil characteristics,

)<sup>2</sup> (4)

, impacted by

), the

two categories: small scale or less than 20 km<sup>2</sup>

mean square error (NRMSE) were applied [40].

*RMSE* <sup>=</sup> <sup>√</sup>

**3.4. Soil moisture comparison**

TDR sensors used in this study are CS616 (CS) from Campbell placed at 2.5, 5, 10, 20 and 30 cm with a latency of every 20 min. The calibration of the CS616 sensors was done according to the manufacturer (ref). In particular, these sensors use linear or quadratic equations to estimate the volumetric water content, depending on the expected range of water content and accuracy requirements. The accuracy reported for these probes is ±2.5 volumetric water content. Measurements of CS sensors are stored in a Campbell CR800 datalogger, which records the data and can then be accessed via peripheral communications using a software interface provided by the company.

FDR sensors tested were Decagon EC-5 and Diviner 2000. Decagon EC-5 (DEC) sensors measure the dielectric constant operates at 70 MHz minimising salinity and texture effects. An advantage is that they provide an accurate sensor reading in almost any soil. Factory calibrations are provided for mineral soils, potting soil and others. The design and measurement frequency allows measurement of volumetric water content (VWC) [37]. The EC-5 sensors were connected by a 3.5-mm stereo jack plug to the Generation I THHINK datalogger collecting data every 20 min [38]. Diviner 2000 [39] is a multi-sensor capacitance probe used to determine soil water content by measuring the frequency change induced by the changing permittivity of the soil permeated by the fringing fields of the capacitor sensor. The probe consists of multiple sensors located at various depths installed in specific access tubes. A high-frequency electric field is created around each sensor (sphere of influence). The sphere of influence is every 10 cm, thus readings are taken in 10 cm depth intervals in the access tube; this allows the sphere of influence for each reading to sample a separate soil horizon. Volumetric soils water measurements are done in real time and the readings are converted to soil moisture using a calibration equation. This universal calibration equation is independent of soil temperature but could be affected by salinity. One advantage is that the access tube is installed with minimum disruption to the soil profile. The accuracy level is better than 99% of the volumetric soil water content (θv) that is taken instantaneously with excellent repeatability. An access tube was allocated at each of the nine test sites into the soil to different depths until 150 cm, and in some cases just above the water table. Readings were registered every 3 days the first weeks and then every 15 days. Results were used applying the calibration equation in order to have volumetric water content and to compare with the gravimetric, TDR, and FDR (Decagon) methods.
