Punjab, Pakistan is famous for rice production in all over the world, but economic indicators are low toward rice contribution in the regional economy. Climatic and physical factors are responsible for rice yield degradation. Suitable land for rice cultivation can be mapped keeping in view these climatic and physical factors. In this research, rice cultivation season was calculated using Moderate Resolution Imaging Spectro-radiometer (MODIS) time series datasets for the complete year 2014. Landsat 8 thermal datasets were obtained for the rice cultivation season and temperature based growth variability maps were generated. The total area under investigation was 13,657 km² out of which 931.61 km² (6.8%) was found to be least suitable, 3316.69 km² (24.2%) was moderately suitable, 6019.63 km² (44%) was highly suitable and 3395.28 km² (24.85%) was not suitable for rice crop cultivation. Results showed that highly suitable area was characterized by a temperature range between 21 and 32 °C, soil pH level between 5.5 and 7.2, soil type was < 78% clay and the soil was imperfectly drained. We compared land suitability map covering the complete land use with rice cultivated area only and found the results as follows: 592 km² (5.9%) rice cultivation was in least suitable, 4385 km² (44%) cultivation was in highly suitable, 2210 km² (23.2%) cultivation was in moderately suitable and 1674 km² (16.8%) cultivation was in not suitable regions. The techniques applied in this research may be used by local farmers to select cropping patterns and land suitability for rice crop.

Net primary production (NPP) is an important indicator of the supply of food and wood. We used a hierarchy model and real time field observations to estimate NPP using satellite imagery. Net radiation received by rice crop canopies was estimated as 27,428 Wm⁻² (215.4 Wm⁻² as averaged) throughout the rice cultivation period (RCP), including 23,168 Wm⁻² (118.3 Wm⁻² as averaged) as shortwave and 4260 Wm⁻² (34.63 Wm⁻² as averaged) as longwave radiation. Soil, sensible and latent heat fluxes were approximated as 3324 Wm⁻², 16,549 Wm⁻², and 7554 Wm⁻², respectively. Water stress on rice crops varied between 0.5838 and 0.1218 from the start until the end of the RCP. Biomass generation declined from 6.09–1.03 g/m² in the tillering and ripening stages, respectively. We added a soil suitability constant (ħα) into the Carnegie-Ames-Stanford Approach (CASA) model to achieve a more precise estimate of yield. Classification results suggest that the total area under rice cultivation was 8861 km². The spatial distribution of rice cultivation as per suitability zone was: 1674 km² was not suitable (NS), 592 km² was less suitable (LS), 2210 km² was moderately suitable (MS) and 4385 km² was highly suitable (HS) soil type with ħα ranges of 0.05–0.25, 0.4–0.6, 0.7–0.75 and 0.85–0.95 of the CASA based yield, respectively. We estimated net production as 1.63 million tons, as per 0.46 ton/ha, 1.2 ton/ha 1.9 ton/ha and 2.4 ton/ha from NS, LS, MS and HS soil types, respectively. The results obtained through this improved CASA model, by addition of the constant ħα, are likely to be useful for agronomists by providing more accurate estimates of NPP.

| Estimation of net crop production before harvest enables agronomists and decision makers to determine the volume of grain precisely. Yield estimation is one of the challenging tasks which is significant to evaluate accurately for farmers. This research was conducted in eastern Punjab Pakistan by incorporating yield/area as reported by Crop Reporting Service Department along with open source satellite datasets. We downloaded three images of each year (2008-2018) from Geological Survey of United States and applied geometric corrections. All the spectral wavelengths were transformed to Top of Atmosphere reflectance and processed the bands in infrared and red wavelengths to generate Ratio Vegetation Index (RVI) and Normalized Difference Vegetation Index (NDVI) datasets. The annual NDVI and RVI based rice yields were compared with CRS based yield records by applying linear regression and generated yield equations. Rice area was estimated using satellite datasets of 11-years by applying supervised classification to 15 bands composite. The satellite extracted areas under rice cultivation, were compared with CRS reported areas by applying linear regression and generated the regression equation. The NDVI of rice crop in 2018 was 0.72 and the predicted yield was estimated as 2.05 ton/ha. Satellite derived rice area in 2018 was 689580 ha, which was substituted in the regression equation to predict the CRS based area that was 654966 ha. The net rice production for the year 2019, was predicted as 1.42 m tons. The remote sensing tools, datasets and the methodology is easy to understand and apply throughout the world to estimate the net productions precisely.